Cube-Cut: Vertebral Body Segmentation in MRI-Data through Cubic-Shaped Divergences

In this article, we present a graph-based method using a cubic template for volumetric segmentation of vertebrae in magnetic resonance imaging (MRI) acquisitions. The user can define the degree of deviation from a regular cube via a smoothness value Δ. The Cube-Cut algorithm generates a directed graph with two terminal nodes (s-t-network), where the nodes of the graph correspond to a cubic-shaped subset of the image’s voxels. The weightings of the graph’s terminal edges, which connect every node with a virtual source s or a virtual sink t, represent the affinity of a voxel to the vertebra (source) and to the background (sink). Furthermore, a set of infinite weighted and non-terminal edges implements the smoothness term. After graph construction, a minimal s-t-cut is calculated within polynomial computation time, which splits the nodes into two disjoint units. Subsequently, the segmentation result is determined out of the source-set. A quantitative evaluation of a C++ implementation of the algorithm resulted in an average Dice Similarity Coefficient (DSC) of 81.33% and a running time of less than a minute.     

Multi-Modal Glioblastoma Segmentation: Man versus Machine

Background and Purpose

Reproducible segmentation of brain tumors on magnetic resonance images is an important clinical need. This study was designed to evaluate the reliability of a novel fully automated segmentation tool for brain tumor image analysis in comparison to manually defined tumor segmentations.

Methods

We prospectively evaluated preoperative MR Images from 25 glioblastoma patients. Two independent expert raters performed manual segmentations. Automatic segmentations were performed using the Brain Tumor Image Analysis software (BraTumIA). In order to study the different tumor compartments, the complete tumor volume TV (enhancing part plus non-enhancing part plus necrotic core of the tumor), the TV+ (TV plus edema) and the contrast enhancing tumor volume CETV were identified. We quantified the overlap between manual and automated segmentation by calculation of diameter measurements as well as the Dice coefficients, the positive predictive values, sensitivity, relative volume error and absolute volume error.

Results

Comparison of automated versus manual extraction of 2-dimensional diameter measurements showed no significant difference (p = 0.29). Comparison of automated versus manual segmentation of volumetric segmentations showed significant differences for TV+ and TV (p<0.05) but no significant differences for CETV (p>0.05) with regard to the Dice overlap coefficients. Spearman''s rank correlation coefficients (ρ) of TV+, TV and CETV showed highly significant correlations between automatic and manual segmentations. Tumor localization did not influence the accuracy of segmentation.

Conclusions

In summary, we demonstrated that BraTumIA supports radiologists and clinicians by providing accurate measures of cross-sectional diameter-based tumor extensions. The automated volume measurements were comparable to manual tumor delineation for CETV tumor volumes, and outperformed inter-rater variability for overlap and sensitivity.     

BRAF Activation Initiates but Does Not Maintain Invasive Prostate Adenocarcinoma

Prostate cancer is the second leading cause of cancer-related deaths in men. Activation of MAP kinase signaling pathway has been implicated in advanced and androgen-independent prostate cancers, although formal genetic proof has been lacking. In the course of modeling malignant melanoma in a tyrosinase promoter transgenic system, we developed a genetically-engineered mouse (GEM) model of invasive prostate cancers, whereby an activating mutation of BRAF^V600E–a mutation found in ∼10% of human prostate tumors–was targeted to the epithelial compartment of the prostate gland on the background of Ink4a/Arf deficiency. These GEM mice developed prostate gland hyperplasia with progression to rapidly growing invasive adenocarcinoma without evidence of AKT activation, providing genetic proof that activation of MAP kinase signaling is sufficient to drive prostate tumorigenesis. Importantly, genetic extinction of BRAF^V600E in established prostate tumors did not lead to tumor regression, indicating that while sufficient to initiate development of invasive prostate adenocarcinoma, BRAF^V600E is not required for its maintenance.     

Real-Time Automatic Segmentation of Optical Coherence Tomography Volume Data of the Macular Region

Optical coherence tomography (OCT) is a high speed, high resolution and non-invasive imaging modality that enables the capturing of the 3D structure of the retina. The fast and automatic analysis of 3D volume OCT data is crucial taking into account the increased amount of patient-specific 3D imaging data. In this work, we have developed an automatic algorithm, OCTRIMA 3D (OCT Retinal IMage Analysis 3D), that could segment OCT volume data in the macular region fast and accurately. The proposed method is implemented using the shortest-path based graph search, which detects the retinal boundaries by searching the shortest-path between two end nodes using Dijkstra’s algorithm. Additional techniques, such as inter-frame flattening, inter-frame search region refinement, masking and biasing were introduced to exploit the spatial dependency between adjacent frames for the reduction of the processing time. Our segmentation algorithm was evaluated by comparing with the manual labelings and three state of the art graph-based segmentation methods. The processing time for the whole OCT volume of 496×644×51 voxels (captured by Spectralis SD-OCT) was 26.15 seconds which is at least a 2-8-fold increase in speed compared to other, similar reference algorithms used in the comparisons. The average unsigned error was about 1 pixel (∼ 4 microns), which was also lower compared to the reference algorithms. We believe that OCTRIMA 3D is a leap forward towards achieving reliable, real-time analysis of 3D OCT retinal data.     

Pituitary Adenoma Volumetry with 3D Slicer

In this study, we present pituitary adenoma volumetry using the free and open source medical image computing platform for biomedical research: (3D) Slicer. Volumetric changes in cerebral pathologies like pituitary adenomas are a critical factor in treatment decisions by physicians and in general the volume is acquired manually. Therefore, manual slice-by-slice segmentations in magnetic resonance imaging (MRI) data, which have been obtained at regular intervals, are performed. In contrast to this manual time consuming slice-by-slice segmentation process Slicer is an alternative which can be significantly faster and less user intensive. In this contribution, we compare pure manual segmentations of ten pituitary adenomas with semi-automatic segmentations under Slicer. Thus, physicians drew the boundaries completely manually on a slice-by-slice basis and performed a Slicer-enhanced segmentation using the competitive region-growing based module of Slicer named GrowCut. Results showed that the time and user effort required for GrowCut-based segmentations were on average about thirty percent less than the pure manual segmentations. Furthermore, we calculated the Dice Similarity Coefficient (DSC) between the manual and the Slicer-based segmentations to proof that the two are comparable yielding an average DSC of 81.97±3.39%.     

Fully Automated Segmentation of the Pons and Midbrain Using Human T1 MR Brain Images

Purpose

This paper describes a novel method to automatically segment the human brainstem into midbrain and pons, called LABS: Landmark-based Automated Brainstem Segmentation. LABS processes high-resolution structural magnetic resonance images (MRIs) according to a revised landmark-based approach integrated with a thresholding method, without manual interaction.

Methods

This method was first tested on morphological T1-weighted MRIs of 30 healthy subjects. Its reliability was further confirmed by including neurological patients (with Alzheimer''s Disease) from the ADNI repository, in whom the presence of volumetric loss within the brainstem had been previously described. Segmentation accuracies were evaluated against expert-drawn manual delineation. To evaluate the quality of LABS segmentation we used volumetric, spatial overlap and distance-based metrics.

Results

The comparison between the quantitative measurements provided by LABS against manual segmentations revealed excellent results in healthy controls when considering either the midbrain (DICE measures higher that 0.9; Volume ratio around 1 and Hausdorff distance around 3) or the pons (DICE measures around 0.93; Volume ratio ranging 1.024–1.05 and Hausdorff distance around 2). Similar performances were detected for AD patients considering segmentation of the pons (DICE measures higher that 0.93; Volume ratio ranging from 0.97–0.98 and Hausdorff distance ranging 1.07–1.33), while LABS performed lower for the midbrain (DICE measures ranging 0.86–0.88; Volume ratio around 0.95 and Hausdorff distance ranging 1.71–2.15).

Conclusions

Our study represents the first attempt to validate a new fully automated method for in vivo segmentation of two anatomically complex brainstem subregions. We retain that our method might represent a useful tool for future applications in clinical practice.     

The Impact of Flap Creation Methods for Sub-Bowman’s Keratomileusis (SBK) on the Central Thickness of Bowman’s Layer

Purpose

To determine the impact of flap creation methods for sub-Bowman’s keratomileusis (SBK) on central Bowman’s layer thickness.

Methods

SBK flaps were made by Moria microkeratome for 20 subjects and by femtosecond (FEMTO) laser for 21 subjects. Corneal sublayer thicknesses were measured by ultra-high resolution optical coherence tomography before SBK and at 1 day, 1 week, 2 weeks, and 1 month afterwards. Each subject was imaged twice on each visit. Thicknesses of central epithelium, Bowman’s layer, flap, and total cornea were calculated using a custom-made automated image processing algorithm. The repeatability of sublayer thickness measurements was tested by the intraclass correlation coefficient (ICC) and by the coefficient of repeatability (CoR) at 1 week post-SBK.

Results

ICCs of the Moria and FEMTO groups were ≥0.959 and ≥0.961 respectively for all sublayer measurements. The segmentation CoRs were less than 6.78% and 5.63% respectively. For both groups, microdistortions were present in the epithelium and Bowman’s layer after SKB. The flap thickness of the Moria group was 9.8 μm (95% confidence interval: 4.8 – 14.8μm) thinner than the FEMTO group one day after SBK (independent samples t-test, P < 0.05). Bowman’s layer became thicker by 1.6 ± 1.1 μm and 1.7 ± 1.6 μm one day post-SBK for the Moria and FEMTO groups (repeated ANOVA, P < 0.05) and then remained stable. Corneal and sublayer thickness were similar between the two groups.

Conclusions

Central Bowman’s layer thickness increased 1 day post-SBK. Flap creation by Moria microkeratome and femtosecond laser did not have significantly different impacts on Bowman’s layer thickness following SBK.

Trial Registration

Chinese Clinical Trial Registry (ChiCTR) NO: ChiCTR-OCH-14004525     

A common perceptual temporal limit of binding synchronous inputs across different sensory attributes and modalities

The human brain processes different aspects of the surrounding environment through multiple sensory modalities, and each modality can be subdivided into multiple attribute-specific channels. When the brain rebinds sensory content information (‘what’) across different channels, temporal coincidence (‘when’) along with spatial coincidence (‘where’) provides a critical clue. It however remains unknown whether neural mechanisms for binding synchronous attributes are specific to each attribute combination, or universal and central. In human psychophysical experiments, we examined how combinations of visual, auditory and tactile attributes affect the temporal frequency limit of synchrony-based binding. The results indicated that the upper limits of cross-attribute binding were lower than those of within-attribute binding, and surprisingly similar for any combination of visual, auditory and tactile attributes (2–3 Hz). They are unlikely to be the limits for judging synchrony, since the temporal limit of a cross-attribute synchrony judgement was higher and varied with the modality combination (4–9 Hz). These findings suggest that cross-attribute temporal binding is mediated by a slow central process that combines separately processed ‘what’ and ‘when’ properties of a single event. While the synchrony performance reflects temporal bottlenecks existing in ‘when’ processing, the binding performance reflects the central temporal limit of integrating ‘when’ and ‘what’ properties.     

Active Contours Driven by Multi-Feature Gaussian Distribution Fitting Energy with Application to Vessel Segmentation

Active contour models are of great importance for image segmentation and can extract smooth and closed boundary contours of the desired objects with promising results. However, they cannot work well in the presence of intensity inhomogeneity. Hence, a novel region-based active contour model is proposed by taking image intensities and ‘vesselness values’ from local phase-based vesselness enhancement into account simultaneously to define a novel multi-feature Gaussian distribution fitting energy in this paper. This energy is then incorporated into a level set formulation with a regularization term for accurate segmentations. Experimental results based on publicly available STructured Analysis of the Retina (STARE) demonstrate our model is more accurate than some existing typical methods and can successfully segment most small vessels with varying width.     

Anti-IL-20 Monoclonal Antibody Suppresses Prostate Cancer Growth and Bone Osteolysis in Murine Models

Interleukin (IL)-20 is a proinflammatory cytokine in the IL–10 family. IL–20 is associated with tumor promotion in the pathogenesis of oral, bladder, and breast cancer. However, little is known about the role of IL–20 in prostate cancer. We hypothesize that IL–20 promotes the growth of prostate cancer cells. Immunohistochemical staining showed that IL–20 and its receptors were expressed in human PC–3 and LNCaP prostate cancer cell lines and in prostate tumor tissue from 40 patients. In vitro, IL–20 upregulated N-cadherin, STAT3, vimentin, fibronectin, RANKL, cathepsin G, and cathepsin K, and increased the migration and colony formation of prostate cancer cells via activated p38, ERK1/2, AKT, and NF-κB signals in PC–3 cells. We investigated the effects of anti-IL–20 monoclonal antibody 7E on prostate tumor growth in vivo using SCID mouse subcutaneous and intratibial xenograft tumor models. In vivo, 7E reduced tumor growth, suppressed tumor-mediated osteolysis, and protected bone mineral density after intratibial injection of prostate cancer cells. We conclude that IL–20 is involved in the cell migration, colony formation, and tumor-induced osteolysis of prostate cancer. Therefore, IL–20 might be a novel target for treating prostate cancer.     

Segmentation of microscopic cell scenes

Different methods for the automated segmentation of microscopic cell scenes are presented with examples. The techniques discussed include edge detection by thresholding, "blob" detection by split-and-merge algorithm, global thresholding using gray-level histograms, hierarchic thresholding using color information, global thresholding using two-dimensional histograms and segmentation by "blob" labeling. Methods are more robust against insignificant changes in the scene and perform more reliably as more a priori knowledge about the scene is incorporated in the segmentation algorithm. The inclusion of both photometric and geometric a priori knowledge can result in a high level of correct segmentations, the cost of which is increased computation time.     

Amino acid–base interactions: a three-dimensional analysis of protein–DNA interactions at an atomic level

To assess whether there are universal rules that govern amino acid–base recognition, we investigate hydrogen bonds, van der Waals contacts and water-mediated bonds in 129 protein–DNA complex structures. DNA–backbone interactions are the most numerous, providing stability rather than specificity. For base interactions, there are significant base–amino acid type correlations, which can be rationalised by considering the stereochemistry of protein side chains and the base edges exposed in the DNA structure. Nearly two-thirds of the direct read-out of DNA sequences involves complex networks of hydrogen bonds, which enhance specificity. Two-thirds of all protein–DNA interactions comprise van der Waals contacts, compared to about one-sixth each of hydrogen and water-mediated bonds. This highlights the central importance of these contacts for complex formation, which have previously been relegated to a secondary role. Although common, water-mediated bonds are usually non-specific, acting as space-fillers at the protein–DNA interface. In conclusion, the majority of amino acid–base interactions observed follow general principles that apply across all protein–DNA complexes, although there are individual exceptions. Therefore, we distinguish between interactions whose specificities are ‘universal’ and ‘context-dependent’. An interactive Web-based atlas of side chain–base contacts provides access to the collected data, including analyses and visualisation of the three-dimensional geometry of the interactions.     

Toxicity outcome in patients treated with modulated arc radiotherapy for localized prostate cancer

Aim

This study evaluates the acute toxicity outcome in patients treated with RapidArc for localized prostate cancer.

Background

Modern technologies allow the delivery of high doses to the prostate while lowering the dose to the neighbouring organs at risk. Whether this dosimetric advantage translates into clinical benefit is not well known.

Materials and methods

Between December 2009 and May 2012, 45 patients with primary prostate adenocarcinoma were treated using RapidArc. All patients received 1.8 Gy per fraction, the median dose to the prostate gland, seminal vesicles, pelvic lymph nodes and surgical bed was 80 Gy (range, 77.4–81 Gy), 50.4 Gy, 50.4 Gy and 77.4 Gy (range, 75.6–79.2 Gy), respectively.

Results

The time between the last session and the last treatment follow up was a median of 10 months (range, 3–24 months). The incidence of grade 3 acute gastrointestinal (GI) and genitourinary (GU) toxicity was 2.2% and 15.5%, respectively. Grade 2 acute GI and GU toxicity occurred in 30% and 27% of patients, respectively. No grade 4 acute GI and GU toxicity were observed. Older patients (>median) or patients with V60 higher than 35% had significantly higher rates of grade ≥2 acute GI toxicity compared with the younger ones.

Conclusions

RapidArc in the treatment of localized prostate cancer is tolerated well with no Grade >3 GI and GU toxicities. Older patients or patients with higher V60 had significantly higher rates of grade ≥2 acute GI toxicity. Further research is necessary to assess definitive late toxicity and tumour control outcome.     

3D Reconstruction of Coronary Artery Vascular Smooth Muscle Cells

Aims

The 3D geometry of individual vascular smooth muscle cells (VSMCs), which are essential for understanding the mechanical function of blood vessels, are currently not available. This paper introduces a new 3D segmentation algorithm to determine VSMC morphology and orientation.

Methods and Results

A total of 112 VSMCs from six porcine coronary arteries were used in the analysis. A 3D semi-automatic segmentation method was developed to reconstruct individual VSMCs from cell clumps as well as to extract the 3D geometry of VSMCs. A new edge blocking model was introduced to recognize cell boundary while an edge growing was developed for optimal interpolation and edge verification. The proposed methods were designed based on Region of Interest (ROI) selected by user and interactive responses of limited key edges. Enhanced cell boundary features were used to construct the cell’s initial boundary for further edge growing. A unified framework of morphological parameters (dimensions and orientations) was proposed for the 3D volume data. Virtual phantom was designed to validate the tilt angle measurements, while other parameters extracted from 3D segmentations were compared with manual measurements to assess the accuracy of the algorithm. The length, width and thickness of VSMCs were 62.9±14.9μm, 4.6±0.6μm and 6.2±1.8μm (mean±SD). In longitudinal-circumferential plane of blood vessel, VSMCs align off the circumferential direction with two mean angles of -19.4±9.3° and 10.9±4.7°, while an out-of-plane angle (i.e., radial tilt angle) was found to be 8±7.6° with median as 5.7°.

Conclusions

A 3D segmentation algorithm was developed to reconstruct individual VSMCs of blood vessel walls based on optical image stacks. The results were validated by a virtual phantom and manual measurement. The obtained 3D geometries can be utilized in mathematical models and leads a better understanding of vascular mechanical properties and function.     

Immediate Risk for Cardiovascular Events and Suicide Following a Prostate Cancer Diagnosis: Prospective Cohort Study

Background

Stressful life events have been shown to be associated with altered risk of various health consequences. The aim of the present study was to investigate whether the emotional stress evoked by a prostate cancer diagnosis increases the immediate risks of cardiovascular events and suicide.

Methods and Findings

We conducted a prospective cohort study by following all men in Sweden who were 30 y or older (n = 4,305,358) for a diagnosis of prostate cancer (n = 168,584) and their subsequent occurrence of cardiovascular events and suicide between January 1, 1961 and December 31, 2004. We used Poisson regression models to calculate relative risks (RRs) and 95% confidence intervals (CIs) of cardiovascular events and suicide among men who had prostate cancer diagnosed within 1 y to men without any cancer diagnosis. The risks of cardiovascular events and suicide were elevated during the first year after prostate cancer diagnosis, particularly during the first week. Before 1987, the RR of fatal cardiovascular events was 11.2 (95% CI 10.4–12.1) during the first week and 1.9 (95% CI 1.9–2.0) during the first year after diagnosis. From 1987, the RR for cardiovascular events, nonfatal and fatal combined, was 2.8 (95% CI 2.5–3.2) during the first week and 1.3 (95% CI 1.3–1.3) during the first year after diagnosis. While the RR of cardiovascular events declined, the RR of suicide was stable over the entire study period: 8.4 (95% CI 1.9–22.7) during the first week and 2.6 (95% CI 2.1–3.0) during the first year after diagnosis. Men 54 y or younger at cancer diagnosis demonstrated the highest RRs of both cardiovascular events and suicide. A limitation of the present study is the lack of tumor stage data, which precluded possibilities of investigating the potential impact of the disease severity on the relationship between a recent diagnosis of prostate cancer and the risks of cardiovascular events and suicide. In addition, we cannot exclude residual confounding as a possible explanation.

Conclusions

Men newly diagnosed with prostate cancer are at increased risks for cardiovascular events and suicide. Future studies with detailed disease characteristic data are warranted. Please see later in the article for the Editors'' Summary     

VolPy: Automated and scalable analysis pipelines for voltage imaging datasets

Voltage imaging enables monitoring neural activity at sub-millisecond and sub-cellular scale, unlocking the study of subthreshold activity, synchrony, and network dynamics with unprecedented spatio-temporal resolution. However, high data rates (>800MB/s) and low signal-to-noise ratios create bottlenecks for analyzing such datasets. Here we present VolPy, an automated and scalable pipeline to pre-process voltage imaging datasets. VolPy features motion correction, memory mapping, automated segmentation, denoising and spike extraction, all built on a highly parallelizable, modular, and extensible framework optimized for memory and speed. To aid automated segmentation, we introduce a corpus of 24 manually annotated datasets from different preparations, brain areas and voltage indicators. We benchmark VolPy against ground truth segmentation, simulations and electrophysiology recordings, and we compare its performance with existing algorithms in detecting spikes. Our results indicate that VolPy’s performance in spike extraction and scalability are state-of-the-art.     

Automatic Nuclei Segmentation in H&E Stained Breast Cancer Histopathology Images

The introduction of fast digital slide scanners that provide whole slide images has led to a revival of interest in image analysis applications in pathology. Segmentation of cells and nuclei is an important first step towards automatic analysis of digitized microscopy images. We therefore developed an automated nuclei segmentation method that works with hematoxylin and eosin (H&E) stained breast cancer histopathology images, which represent regions of whole digital slides. The procedure can be divided into four main steps: 1) pre-processing with color unmixing and morphological operators, 2) marker-controlled watershed segmentation at multiple scales and with different markers, 3) post-processing for rejection of false regions and 4) merging of the results from multiple scales. The procedure was developed on a set of 21 breast cancer cases (subset A) and tested on a separate validation set of 18 cases (subset B). The evaluation was done in terms of both detection accuracy (sensitivity and positive predictive value) and segmentation accuracy (Dice coefficient). The mean estimated sensitivity for subset A was 0.875 (±0.092) and for subset B 0.853 (±0.077). The mean estimated positive predictive value was 0.904 (±0.075) and 0.886 (±0.069) for subsets A and B, respectively. For both subsets, the distribution of the Dice coefficients had a high peak around 0.9, with the vast majority of segmentations having values larger than 0.8.     

Automated segmentation of routinely hematoxylin-eosin-stained microscopic images by combining support vector machine clustering and active contour models

OBJECTIVE: To develop a method for the automated segmentation of images of routinely hematoxylin-eosin (H-E)-stained microscopic sections to guarantee correct results in computer-assisted microscopy. STUDY DESIGN: Clinical material was composed 50 H-E-stained biopsies of astrocytomas and 50 H-E-stained biopsies of urinary bladder cancer. The basic idea was to use a support vector machine clustering (SVMC) algorithm to provide gross segmentation of regions holding nuclei and subsequently to refine nuclear boundary detection with active contours. The initialization coordinates of the active contour model were defined using a SVMC pixel-based classification algorithm that discriminated nuclear regions from the surrounding tissue. Starting from the boundaries of these regions, the snake fired and propagated until converging to nuclear boundaries. RESULTS: The method was validated for 2 different types of H-E-stained images. Results were evaluated by 2 histopathologists. On average, 94% of nuclei were correctly delineated. CONCLUSION: The proposed algorithm could be of value in computer-based systems for automated interpretation of microscopic images.     

Circulating Thyroxine,Thyroid-Stimulating Hormone,and Hypothyroid Status and the Risk of Prostate Cancer

Background

Thyroid hormones may influence risk of cancer through their role in cell differentiation, growth, and metabolism. One study of circulating thyroid hormones supports this hypothesis with respect to prostate cancer. We undertook a prospective analysis of thyroid hormones and prostate cancer risk in the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study.

Methods

Within the ATBC Study, a randomized controlled trial of α-tocopherol and β-carotene supplements and cancer incidence in male smokers, 402 prostate cancer cases were sampled. Controls were matched 2∶1 to cases on age and date of blood collection. Odds ratios (OR) and 95% confidence intervals (CI) of prostate cancer were estimated for quintiles of serum total and free thyroxine (T4), thyroid-stimulating hormone (TSH), thyroid-binding globulin (TBG), and by categories of thyroid status.

Results

Men with serum higher TSH had a decreased risk of prostate cancer compared to men with lower TSH (Q5 vs. Q1–4: OR = 0.70, 95% CI: 0.51–0.97, p = 0.03). When the T4 and TSH measurements were combined to define men as hypothyroid, euthyroid or hyperthyroid, hypothyroid men had a lower risk of prostate cancer compared to euthyroid men (OR = 0.48, 95% CI = 0.28–0.81, p = 0.006). We observed no association between hyperthyroid status and risk, although the number of hyperthyroid men with prostate cancer was small (n = 9).

Conclusions

In this prospective study of smokers, men with elevated TSH and those classified as being in a hypothyroid state were at decreased risk of prostate cancer. Future studies should examine the association in other populations, particularly non-smokers and other racial/ethnic groups.