Three-dimensional registration of prostate images from histology and ultrasound

Taylor, Lawrence S.; Porter, Brian; Nadasdy, Gyongyi; Sant'Agnese, Paul A. di; Pasternack, D.; Wu, Zhe; Baggs, Raymond B.; Rubens, Deborah J.; Parker, Kevin J.

doi:10.1016/j.ultrasmedbio.2003.10.005

Cited by 38 publications

(30 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Volume to volume registration was not possible due to the fact that for each 4 mm-thick MRI slice there was only one 20 microns-thick histology section. A 3D surface registration, which has been done before for prostate US images and histology sections [25], was a possibility. However, the main benefit of this approach would be an easier assignment of corresponding slices/sections, with negligible impact on the accuracy of the MRI to histology registration.…”

Section: Discussionmentioning

confidence: 99%

In vivo 3T and ex vivo 7T diffusion tensor imaging of prostate cancer: Correlation with histology

Uribe

Jones

Chang

et al. 2015

Magnetic Resonance Imaging

View full text Add to dashboard Cite

The purpose of this work was to test whether fractional anisotropy (FA) can contribute to the diagnosis and grading of prostate cancer. Turbo spin echo T2-weighted (T2W) and single shot echo planar imaging diffusion tensor imaging (EPI DTI) data were collected from 13 subjects with biopsy proven prostate cancer prior to surgical removal of the gland. Rapid acquisition with relaxation enhancement (RARE) T2W and spin-echo DTI data were acquired ex-vivo from the fixed prostatectomy specimens. Digitized whole mount histology sections, examined and annotated by a pathologist, were registered to the in-vivo and ex-vivo DTI data, and the average values of apparent diffusion coefficient (ADC) and FA were calculated from ROIs encompassing normal and cancerous peripheral zone (PZ). In addition, Monte Carlo simulations were carried out to assess the dependence of the apparent FA on the ADC values for different signal to noise ratios (SNRs). ADC values were significantly lower in tumors than in normal PZ both in-vivo and ex-vivo, while the difference in FA values between tumors and normal PZ was significant only in-vivo. Paired t-test showed significant difference between in-vivo and ex-vivo FA values in tumors, but not in the normal PZ. The simulations showed that lower SNR results in an increasing overestimation of the FA values with decreasing ADC. These results suggest that the in-vivo increase in FA values in tumors is due to low SNR, rather than the presence of cancer. The results of this study suggest that FA does not contribute significantly to the diagnostic capabilities of DTI in prostate cancer.

show abstract

Section: Discussionmentioning

confidence: 99%

In vivo 3T and ex vivo 7T diffusion tensor imaging of prostate cancer: Correlation with histology

Uribe

Jones

Chang

et al. 2015

Magnetic Resonance Imaging

View full text Add to dashboard Cite

show abstract

“…Registration accuracy was assessed by BP by measurement of urethral offset, and also by an overlap metric of intersection over union, applied to the whole gland (20). Fusion was achieved with an in-house correlative (21) program, and images viewed in IRIS Explorer (Numerical Algorithms Group, Downers Grove, IL), as described previously (21,22). The best 3D correlation maps the gray scale B-mode data into the histology frame of reference.…”

Section: Reference Standardmentioning

confidence: 99%

Prostate Cancer: Three-dimensional Sonoelastography for in Vitro Detection

Taylor¹,

Rubens²,

Porter³

et al. 2005

Radiology

Self Cite

View full text Add to dashboard Cite

PURPOSE: To prospectively evaluate the accuracy of 3D sonoelastography for detection of prostate cancer relative to gray scale sonography in vitro.METHODS: Using an Institutional Review Board-approved, HIPAA compliant protocol with informal consent, 19 prostatectomy specimens from patients 46 to 70 years of age with biopsy proven prostate cancer were scanned in 3D using conventional B-scan and sonoelastography using vibrations above 100Hz.Step-sectioned whole-mount histology was utilized to create a 3D volume of the prostate and tumors within it. B-scan ultrasound images and regions of low vibration in the sonoelastography images (hard regions) were formatted in 3D. The lesions in the nineteen cases were analyzed as two groups: G1) pathology-confirmed tumors of 1.0 cc or greater; and G2) pathology-confirmed tumor size less than 1.0 cc. G1 cases were evaluated for B-scan ultrasound and sonoelastography vs. histology as a reference standard and were scored as either a True Positive, a False Positive, a True Negative, or a False Negative. G2 cases were evaluated for sonoelastography only. True positives required 3D lesion correlation between pathology and imaging data. Conventional definitions of accuracy and sensitivity were employed to calculate these statistics. RESULTS: G1 (7 lesions with tumor volume 1.0 cc or greater): Sonoelastography: accuracy of 55%, sensitivity of 71%. B-scan: accuracy of 17%, sensitivity of 29%. Mean tumor size is 3.1cc +/-2.1cc. G2 (22 lesions with tumor volume less than 1.0 cc). Mean tumor size is 0.32 cc +/-0.21 cc. Sonoelastography: accuracy of 34%, sensitivity of 41%, false positives: 6.CONCLUSIONS: Sonoelastography performed considerably better than gray scale sonography in the detection of prostate cancer tumors over 1 cc.

show abstract

“…However, some 1018 Ultrasound in Medicine and Biology Volume 33, Number 7, 2007 researchers have developed more accurate approaches to solve this problem. For example see Taylor et al (2004) who have created fiducial points in the tissue using a four prong device to register their in vitro ultrasound images to histopathology. They have used a 3D correlation registration algorithm and reported accuracies between 89% and 94%.…”

Section: Fundamentals Of Tissue Characterization Based On Ultrasound mentioning

confidence: 99%

Computer-Aided Diagnosis of Prostate Cancer with Emphasis on Ultrasound-Based Approaches: A Review

Moradi

Mousavi

Abolmaesumi

2007

Ultrasound in Medicine & Biology

View full text Add to dashboard Cite

Three-dimensional registration of prostate images from histology and ultrasound

Cited by 38 publications

References 11 publications

In vivo 3T and ex vivo 7T diffusion tensor imaging of prostate cancer: Correlation with histology

In vivo 3T and ex vivo 7T diffusion tensor imaging of prostate cancer: Correlation with histology

Prostate Cancer: Three-dimensional Sonoelastography for in Vitro Detection

Computer-Aided Diagnosis of Prostate Cancer with Emphasis on Ultrasound-Based Approaches: A Review

Contact Info

Product

Resources

About