Search

Three-dimensional (3D) prostate image segmentation is useful for cancer diagnosis and therapy guidance, but can be time-consuming to perform manually and involves varying levels of difficulty and interoperator variability within the prostatic base, midgland (MG), and apex. In this study, the authors measured accuracy and interobserver variability in the segmentation of the prostate on T2-weighted endorectal magnetic resonance (MR) imaging within the whole gland (WG), and separately within the apex, midgland, and base regions.The authors collected MR images from 42 prostate cancer patients. Prostate border delineation was performed manually by one observer on all images and by two other observers on a subset of ten images. The authors used complementary boundary-, region-, and volume-based metrics [mean absolute distance (MAD), Dice similarity coefficient (DSC), recall rate, precision rate, and volume difference (ΔV)] to elucidate the different types of segmentation errors that they observed. Evaluation for expert manual and semiautomatic segmentation approaches was carried out. Compared to manual segmentation, the authors' semiautomatic approach reduces the necessary user interaction by only requiring an indication of the anteroposterior orientation of the prostate and the selection of prostate center points on the apex, base, and midgland slices. Based on these inputs, the algorithm identifies candidate prostate boundary points using learned boundary appearance characteristics and performs regularization based on learned prostate shape information.The semiautomated algorithm required an average of 30 s of user interaction time (measured for nine operators) for each 3D prostate segmentation. The authors compared the segmentations from this method to manual segmentations in a single-operator (mean whole gland MAD = 2.0 mm, DSC = 82%, recall = 77%, precision = 88%, and ΔV = - 4.6 cm(3)) and multioperator study (mean whole gland MAD = 2.2 mm, DSC = 77%, recall = 72%, precision = 86%, and ΔV = - 4.0 cm(3)). These results compared favorably with observed differences between manual segmentations and a simultaneous truth and performance level estimation reference for this data set (whole gland differences as high as MAD = 3.1 mm, DSC = 78%, recall = 66%, precision = 77%, and ΔV = 15.5 cm(3)). The authors found that overall, midgland segmentation was more accurate and repeatable than the segmentation of the apex and base, with the base posing the greatest challenge.The main conclusions of this study were that (1) the semiautomated approach reduced interobserver segmentation variability; (2) the segmentation accuracy of the semiautomated approach, as well as the accuracies of recently published methods from other groups, were within the range of observed expert variability in manual prostate segmentation; and (3) further efforts in the development of computer-assisted segmentation would be most productive if focused on improvement of segmentation accuracy and reduction of variability within the prostatic apex and base.

This paper presents the results of an in vivo clinical study to accurately characterize prostate cancer using new features of ultrasound RF time series.The mean central frequency and wavelet features of ultrasound RF time series from seven patients are used along with an elaborate framework of ultrasound to histology registration to identify and verify cancer in prostate tissue regions as small as 1.7 mm x 1.7 mm.In a leave-one-patient-out cross-validation strategy, an average classification accuracy of 76% and the area under ROC curve of 0.83 are achieved using two proposed RF time series features. The results statistically significantly outperform those achieved by previously reported features in the literature. The proposed features show the clinical relevance of RF time series for in vivo characterization of cancer.

Determining the intra-prostatic spatial distribution and grade of prostate cancer before treatment may support improved diagnosis, therapy selection, or guidance of intra-prostatic lesion-focused therapies (e.g., radiation boosting or ablative focal therapy). Several in vivo imaging modalities are showing promise for imaging the intra-prostatic distribution of cancer. Evaluations of such imaging modalities ideally include comparisons to registered histological examinations of prostatectomy specimens, the clinically accepted “gold standard” for staging and grading prostate cancer. The registration of histology to ex vivo magnetic resonance (MR) images supports these challenging in vivo registrations by reconstructing 3D spatial information that is lost during the process of acquiring histology sections. In the work described in this chapter, ex vivo MR and histology images were acquired from nine formalin-fixed radical prostatectomy specimens which had been marked with extrinsic fiducials designed to be visible in these modalities. The histology images were registered retrospectively to the MR images using a novel algorithm based on the minimization of fiducial registration error between fiducial cross-sections on histology images and parametric fiducial curves on MR images. The 3D target registration error (TRE) was quantified based on the post-registration misalignment of manually identified homologous landmarks (3–7 per section, 184 in total), and was compared to two previously developed methods: (1) a method based on the guidance of the coarse slicing of specimens and (2) a method based on additional imaging of the images of the thick tissue slices. The proposed method yielded a mean ± standard deviation TRE of 0.71 ± 0.38 mm, 0.38–0.63 mm lower [95 % confidence interval (CI)] than the image-guided-slicing-based method, and within 0.13 mm (95 % CI) of the tissue-slice-imaging-based method. One component of the proposed method was able to refine the result from the image-guided-slicing-based method to within 0.13 mm (95 % CI) of the proposed method. The proposed method also resulted in a 70 % decrease in specimen processing time compared to the image-guided-slicing-based approach previously implemented at our center.

Background: Guidelines for localizing prostate cancer on imaging are ideally informed by registered post-prostatectomy histology. 3D histology reconstruction methods can support this by reintroducing 3D spatial information lost during histology processing. The need to register small, high-grade foci drives a need for high accuracy. Accurate 3D reconstruction method design is impacted by the answers to the following central questions of this work. (1) How does prostate tissue deform during histology processing? (2) What spatial misalignment of the tissue sections is induced by microtome cutting? (3) How does the choice of reconstruction model affect histology reconstruction accuracy? Materials and Methods: Histology, paraffin block face and magnetic resonance images were acquired for 18 whole mid-gland tissue slices from six prostates. 7-15 homologous landmarks were identified on each image. Tissue deformation due to histology processing was characterized using the target registration error (TRE) after landmark-based registration under four deformation models (rigid, similarity, affine and thin-plate-spline [TPS]). The misalignment of histology sections from the front faces of tissue slices was quantified using manually identified landmarks. The impact of reconstruction models on the TRE after landmark-based reconstruction was measured under eight reconstruction models comprising one of four deformation models with and without constraining histology images to the tissue slice front faces. Results: Isotropic scaling improved the mean TRE by 0.8-1.0 mm (all results reported as 95% confidence intervals), while skew or TPS deformation improved the mean TRE by

Image registration is an important tool for imaging validation studies investigating the effect of underlying focal disease on the imaging signal. The strength of the conclusions drawn from these analyses is limited by statistical power. Based on the observation that in this context, statistical power depends in part on uncertainty arising from registration error, we derive a power calculation formula relating registration error, sample size, and the minimum detectable difference between normal and pathologic regions on imaging. Statistical mappings between target registration error and fractional overlap metrics are also derived, and Monte Carlo simulations are used to evaluate the derived models and test the strength of their assumptions.

Novel imaging modalities are pushing the boundaries of what is possible in medical imaging, but their signal properties are not always well understood. The evaluation of these novel imaging modalities is critical to achieving their research and clinical potential. Image registration of novel modalities to accepted reference standard modalities is an important part of characterizing the modalities and elucidating the effect of underlying focal disease on the imaging signal. The strengths of the conclusions drawn from these analyses are limited by statistical power. Based on the observation that in this context, statistical power depends in part on uncertainty arising from registration error, we derive a power calculation formula relating registration error, number of subjects, and the minimum detectable difference between normal and pathologic regions on imaging, for an imaging validation study design that accommodates signal correlations within image regions. Monte Carlo simulations were used to evaluate the derived models and test the strength of their assumptions, showing that the model yielded predictions of the power, the number of subjects, and the minimum detectable difference of simulated experiments accurate to within a maximum error of 1% when the assumptions of the derivation were met, and characterizing sensitivities of the model to violations of the assumptions. The use of these formulae is illustrated through a calculation of the number of subjects required for a case study, modeled closely after a prostate cancer imaging validation study currently taking place at our institution. The power calculation formulae address three central questions in the design of imaging validation studies: (1) What is the maximum acceptable registration error? (2) How many subjects are needed? (3) What is the minimum detectable difference between normal and pathologic image regions?

Methods for 3D histology reconstruction from sparse 2D digital histology images depend on knowledge about the positions, orientations, and deformations of tissue slices due to the histology process. This work quantitatively evaluates typical assumptions about the position and orientation of whole-mount prostate histology sections within coarsely sliced tissue blocks and about the deformation of tissue during histological processing and sectioning. 3-5 midgland tissue blocks from each of 7 radical prostatectomy specimens were imaged using magnetic resonance imaging before histology processing. After standard whole-mount paraffin processing and sectioning, the resulting sections were digitised. Homologous anatomic landmarks were identified on 22 midgland histology and MR images. Orientations and depths of sections relative to the front faces of the tissue blocks were measured based on the best-fit plane through the landmarks on the MR images. The mean±std section orientation was 1.7±1.1° and the mean±std depth of the sections was 1.0±0.5 mm. Deformation was assessed by using four transformation models (rigid, rigid+scale, affine and thin-plate-spline (TPS)) to align landmarks from histology and MR images, and evaluating each by measuring the target registration error (TRE) using a leave-one-out cross-validation. The rigid transformation model had higher mean TRE (p

The correlation of multiple in vivo imaging modalities with graded whole-mount histology ideally requires co-registration with appropriately quantified accuracy. This work presents a framework for the co-registration of multiple in vivo prostate images with histology as a composition of pair wise registrations. This work also proposes a methodology for the measurement of the accuracy of these registrations based on a composition of target registration errors calculated from homologous landmarks in registered pairs of images. Preliminary results from the implementation of components of the framework are reported.

Purpose: To present and evaluate a method for registration of whole-mount prostate digital histology images to ex vivo magnetic resonance (MR) images. Materials and Methods: Nine radical prostatectomy specimens were marked with 10 strand-shaped fiducial markers per specimen, imaged with T1- and T2-weighted 3T MRI protocols, sliced at 4.4-mm intervals, processed for whole-mount histology, and the resulting histological sections (3–5 per specimen, 34 in total) were digitized. The correspondence between fiducial markers on histology and MR images yielded an initial registration, which was refined by a local optimization technique, yielding the least-squares best-fit affine transformation between corresponding fiducial points on histology and MR images. Accuracy was quantified as the postregistration 3D distance between landmarks (3–7 per section, 184 in total) on histology and MR images, and compared to a previous state-of-the-art registration method. Results: The proposed method and previous method had mean (SD) target registration errors of 0.71 (0.38) mm and 1.21 (0.74) mm, respectively, requiring 3 and 11 hours of processing time, respectively. Conclusion: The proposed method registers digital histology to prostate MR images, yielding 70% reduced processing time and mean accuracy sufficient to achieve 85% overlap on histology and ex vivo MR images for a 0.2 cc spherical tumor. J. Magn. Reson. Imaging 2012; 36:1402–1412. © 2012 Wiley Periodicals, Inc.

Personalized treatment of prostate cancer would be enhanced by an assessment of cancer stage and grade from imaging, the validation of which requires the accurate co-registration of in vivo images with a gold standard for stage and grade established by histopathology. We present a visualization tool supporting an image-guided approach enabling the acquisition of histopathology images parallel to the in vivo imaging planes, simplifying this registration. This tool decreases imaging-to-specimen landmark alignment error by 62%, and decreases the time required to mark the slicing plane on the specimen by 47%. Preliminary results from our method demonstrate the alignment of regions suspicious for cancer on T2w MRI with confirmed cancer foci on histopathology, and we calculate a sub-millimeter in-plane target registration error.

Rodent models of retinal degenerative diseases are used by vision scientists to develop therapies and to understand mechanisms of disease progression. Measurement of changes to the thickness of the various retinal layers provides an objective metric to evaluate the performance of the therapy. Because invasive histology is terminal and provides only a single data point, non-invasive imaging modalities are required to better study progression, and to reduce the number of animals used in research. Optical Coherence Tomography (OCT) has emerged as a dominant imaging modality for human ophthalmic imaging, but has only recently gained significant attention for rodent retinal imaging. OCT provides cross section images of retina with micron-scale resolution which permits measurement of the retinal layer thickness. However, in order to be useful to vision scientists, a significant fraction of the retinal surface needs to be measured. In addition, because the retinal thickness normally varies as a function of distance from optic nerve head, it is critical to sample all regions of the retina in a systematic fashion. We present a longitudinal study of OCT to measure retinal degeneration in rats which have undergone optic nerve axotomy, a well characterized form of rapid retinal degeneration. Volumetric images of the retina acquired with OCT in a time course study were segmented in 2D using a semi-automatic segmentation algorithm. Then, using a 3D algorithm, thickness measurements were quantified across the surface of the retina for all volume segmentations. The resulting maps of the changes to retinal thickness over time represent the progression of degeneration across the surface of the retina during injury. The computational tools complement OCT retinal volumetric acquisition, resulting in a powerful tool for vision scientists working with rodents.

Constructing a one to one correspondence between whole brain MR image scans is a problem of critical importance in neuroimaging analyses. We present a framework to combine the strength of both surface-based and volumetric-based analyses for consistent, bijective data transfer between brain coordinate systems.

Optical Coherence Tomography is a powerful tool for diagnostic imaging of the ocular posterior chamber. Recent advances in OCT technology have facilitated acquisition of high resolution volumetric images of the retina and optic nerve head. In this report, we investigate optic nerve head imaging in humans using a home-built laboratory grade OCT system in the 800nm wavelength region. We also introduce the development of a computational model of the optic nerve head morphology in order to study physiological changes which may be associated with elevated intra-ocular pressure.

volume is critical in treatment planning. Whether by expert guidelines, retrograde urethrography, bony landmark, ultrasound, or CT-based planning, there is discordant anatomic information regarding the distance between apex and the bulb, also known as the genitourinary diaphragm (GUD) or membranous urethra length (MUL). In this study, we use MRI with endorectal coil in a large series of patients to better characterize the MUL in radiation therapy planning. Materials/Methods: Patients with early stage, clinically localized prostate cancer (Gleason 6 and 7, cT1c-cT2b, PSA 85% interobserver agreement within 2 mm measurement. There was no association between MUL and patient age, T-stage, Gleason score, patient height, weight, or NCCN risk group. Conclusions: Although there remains variability, the median MUL, or GUD length, is 15 mm. This report likely represents the largest evaluation of the pre-treatment MUL in men with prostate cancer treated with external beam radiation. Our data challenges traditional guidelines between the apex and bulb from smaller series and adds to critical anatomic information of male urethral length and prostate apex anatomy. Prostate MRI with endorectal coil produces a reliable MUL and should be used to optimize target volume delineation at the prostate apex. Author Disclosure: N.K. Taunk: None. O. Akin: None. X. Pei: None. M. Zelefsky: None.

Purpose: Evaluation ofin vivo prostate imaging modalities for determining the spatial distribution and aggressiveness of prostate cancer ideally requires accurate registration of images to an accepted reference standard, such as histopathological examination of radical prostatectomy specimens. Three-dimensional (3D) reconstruction of prostate histology facilitates these registration-based evaluations by reintroducing 3D spatial information lost during histology processing. Because the reconstruction accuracy may constrain the clinical questions that can be answered with these data, it is important to assess the tradeoffs between minimally disruptive methods based on intrinsic image information and potentially more robust methods based on extrinsic fiducial markers. Methods: Ex vivo magnetic resonance (MR) images and digitized whole-mount histology images from 12 radical prostatectomy specimens were used to evaluate four 3D histology reconstruction algorithms. 3D reconstructions were computed by registering each histology image to the corresponding ex vivo MR image using one of two similarity metrics (mutual information or fiducial registration error) and one of two search domains (affine transformations or a constrained subset thereof). The algorithms were evaluated for accuracy using the mean target registration error (TRE) computed from homologous intrinsic point landmarks (3–16 per histology section; 232 total) identified on histology and MR images, and for the sensitivity of TRE to rotational, translational, and scaling initialization errors. Results: The algorithms using fiducial registration error and mutual information had mean ± standard deviation TREs of 0.7 ± 0.4 and 1.2 ± 0.7 mm, respectively, and one algorithm using fiducial registration error and affine transforms had negligible sensitivities to initialization errors. The postoptimization values of the mutual information-based metric showed evidence of errors due to both the optimizer and the similarity metric, and variation of parameters of the mutual information-based metric did not improve its performance. Conclusions: The extrinsic fiducial-based algorithm had lower mean TRE and lower sensitivity to initialization than the intrinsic intensity-based algorithm using mutual information. A model relating statistical power to registration error for certain imaging validation study designs estimated that a reconstruction algorithm with a mean TRE of 0.7 mm would require 27% fewer subjects than the method used to initialize the algorithms (mean TRE 1.3 ± 0.7 mm), suggesting the choice of reconstruction technique can have a substantial impact on the design of imaging validation studies, and on their overall cost.

Purpose: An accurate determination of stage and grade from in vivo prostate cancerimaging would support improved biopsy guidance, therapy selection and, possibly, focal therapy guidance. Validation of prostate cancerimaging modalities ideally requires accurate registration of in vivo imaging to histopathology, where accurate diagnoses can be made. An important intermediate component of this registration aligns ex vivo imaging to histology. This work describes and evaluates a simplified approach, developed in our laboratory, to the registration of digital prostate histopathology to ex vivo magnetic resonance(MR)images. Our approach obviates three elements typical of existing methods: (1) guidance of specimen slicing, (2) imaging/photography of sliced tissue blocks, and (3) dependence on the appearance of salient structures within the image to guide registration. Methods: Six resected prostate specimens were marked with strand‐shaped fiducials, imaged with T1‐ and T2‐weighted MRI before and after image‐guided specimen slicing, and then processed for histology as whole mount specimens. Direct fiducial‐based registration from histology to whole specimen ex vivo MRI was performed by establishing a correspondence between fiducial cross‐sections on histology and fiducial space curves on whole specimen MRI. The target registration error (TRE) of 93 intrinsic landmarks across 21 histology images was calculated for this method, for an indirect registration using tissue block MRI and for a registration based on image‐guidance of slicing. Results: The image‐guided, indirect and direct registrations yielded TREs of 1.2 mm, 0.8 mm and 0.8 mm respectively. A two‐tail paired t‐test failed to show a significant difference between direct and indirect registration mean TREs (p = 0.8, 95% confidence interval, −0.08 to 0.10 mm), and showed that both TREs were lower than that of the image‐guided registration (p = 0.0001). Conclusions: The direct registration is a suitable replacement for existing, more complex registration methods for aligning histopathology to ex vivo imaging.