Quantitative analysis of multiparametric prostate MR images: differentiation between prostate cancer and normal tissue and correlation with Gleason score--a computer-aided diagnosis development study.
TL;DR: The combination of 10th percentile ADC, average ADC, and T2-weighted skewness with CAD is promising in the differentiation of prostate cancer from normal tissue.
Abstract: The findings of this retrospective study suggest that computer-aided diagnosis based on a combination of quantitative imaging features derived from multiparametric endorectal MR images might be effective in distinguishing prostate cancer from normal prostate tissue and that apparent diffusion coefficient features and volume transfer constant moderately correlate with tumor Gleason score.
Citations
More filters
TL;DR: The diagnostic accuracy of multiparametric magnetic resonance imaging (mpMRI) for the detection of clinically significant prostate cancer (PCa) is concluded that mpMRI is able to detect significant PCa and may used to target prostate biopsies.
680 citations
TL;DR: Several Haralick-based texture features appear useful for prostate cancer detection and GS assessment and several HarAlick texture features may differentiate non-cancerous and cancerous prostate tissue.
Abstract: To investigate Haralick texture analysis of prostate MRI for cancer detection and differentiating Gleason scores (GS). One hundred and forty-seven patients underwent T2- weighted (T2WI) and diffusion-weighted prostate MRI. Cancers ≥0.5 ml and non-cancerous peripheral (PZ) and transition (TZ) zone tissue were identified on T2WI and apparent diffusion coefficient (ADC) maps, using whole-mount pathology as reference. Texture features (Energy, Entropy, Correlation, Homogeneity, Inertia) were extracted and analysed using generalized estimating equations. PZ cancers (n = 143) showed higher Entropy and Inertia and lower Energy, Correlation and Homogeneity compared to non-cancerous tissue on T2WI and ADC maps (p-values: 7: p = 0.0069) and lower Energy (GS 6 vs. 7: p = 0.0116, 6 vs. >7: p = 0.0039). ADC map Energy (p = 0.0102) and Entropy (p = 0.0019) were significantly different in GS ≤3 + 4 versus ≥4 + 3 cancers; ADC map Entropy remained significant after controlling for the median ADC (p = 0.0291). Several Haralick-based texture features appear useful for prostate cancer detection and GS assessment. • Several Haralick texture features may differentiate non-cancerous and cancerous prostate tissue.
• Tumour Energy and Entropy on ADC maps correlate with Gleason score.
• T2w-image-derived texture features are not associated with the Gleason score.
322 citations
TL;DR: This Review critically evaluates the existing literature focused on defining the high-risk population, the management of patients with high- risk prostate cancer, and future directions to optimize care.
Abstract: Approximately 15% of patients with prostate cancer are diagnosed with high-risk disease. However, the current definitions of high-risk prostate cancer include a heterogeneous group of patients with a range of prognoses. Some have the potential to progress to a lethal phenotype that can be fatal, while others can be cured with treatment of the primary tumour alone. The optimal management of this patient subgroup is evolving. A refined classification scheme is needed to enable the early and accurate identification of high-risk disease so that more-effective treatment paradigms can be developed. We discuss several principles established from clinical trials, and highlight other questions that remain unanswered. This Review critically evaluates the existing literature focused on defining the high-risk population, the management of patients with high-risk prostate cancer, and future directions to optimize care.
316 citations
TL;DR: Machine learning-based automatic classification of prostate cancer aggressiveness by combining apparent diffusion coefficient (ADC) and T2-weighted (T2-w) MRI-based texture features together with sample augmentation can help to obtain reasonably accurate classification of Gleason patterns are presented.
Abstract: Noninvasive, radiological image-based detection and stratification of Gleason patterns can impact clinical outcomes, treatment selection, and the determination of disease status at diagnosis without subjecting patients to surgical biopsies. We present machine learning-based automatic classification of prostate cancer aggressiveness by combining apparent diffusion coefficient (ADC) and T2-weighted (T2-w) MRI-based texture features. Our approach achieved reasonably accurate classification of Gleason scores (GS) 6(3 + 3) vs. ≥7 and 7(3 + 4) vs. 7(4 + 3) despite the presence of highly unbalanced samples by using two different sample augmentation techniques followed by feature selection-based classification. Our method distinguished between GS 6(3 + 3) and ≥7 cancers with 93% accuracy for cancers occurring in both peripheral (PZ) and transition (TZ) zones and 92% for cancers occurring in the PZ alone. Our approach distinguished the GS 7(3 + 4) from GS 7(4 + 3) with 92% accuracy for cancers occurring in both the PZ and TZ and with 93% for cancers occurring in the PZ alone. In comparison, a classifier using only the ADC mean achieved a top accuracy of 58% for distinguishing GS 6(3 + 3) vs. GS ≥7 for cancers occurring in PZ and TZ and 63% for cancers occurring in PZ alone. The same classifier achieved an accuracy of 59% for distinguishing GS 7(3 + 4) from GS 7(4 + 3) occurring in the PZ and TZ and 60% for cancers occurring in PZ alone. Separate analysis of the cancers occurring in TZ alone was not performed owing to the limited number of samples. Our results suggest that texture features derived from ADC and T2-w MRI together with sample augmentation can help to obtain reasonably accurate classification of Gleason patterns.
306 citations
TL;DR: 10th percentile ADC correlated with Gleason score better than did other ADC parameters, suggesting that 10th percentile ADC may prove to be optimal for differentiating low-grade from intermediate- or high-grade prostate cancer with diffusion-weighted MR imaging.
Abstract: When apparent diffusion coefficient (ADC) parameters are determined from whole-lesion histogram analysis, the 10th percentile ADC correlates better with the Gleason score and enables more accurate differentiation of prostate cancer lesions with a Gleason score of 6 from those with a Gleason score of at least 7 than do other ADC parameters commonly used in the literature.
254 citations
References
More filters
TL;DR: A representation and interpretation of the area under a receiver operating characteristic (ROC) curve obtained by the "rating" method, or by mathematical predictions based on patient characteristics, is presented and it is shown that in such a setting the area represents the probability that a randomly chosen diseased subject is (correctly) rated or ranked with greater suspicion than a random chosen non-diseased subject.
Abstract: A representation and interpretation of the area under a receiver operating characteristic (ROC) curve obtained by the "rating" method, or by mathematical predictions based on patient characteristics, is presented. It is shown that in such a setting the area represents the probability that a randomly chosen diseased subject is (correctly) rated or ranked with greater suspicion than a randomly chosen non-diseased subject. Moreover, this probability of a correct ranking is the same quantity that is estimated by the already well-studied nonparametric Wilcoxon statistic. These two relationships are exploited to (a) provide rapid closed-form expressions for the approximate magnitude of the sampling variability, i.e., standard error that one uses to accompany the area under a smoothed ROC curve, (b) guide in determining the size of the sample required to provide a sufficiently reliable estimate of this area, and (c) determine how large sample sizes should be to ensure that one can statistically detect difference...
19,398 citations
TL;DR: The reduction in overall cancer death rates since 1990 in men and 1991 in women translates to the avoidance of about 1,024,400 deaths from cancer, which can be accelerated by applying existing cancer control knowledge across all segments of the population, with an emphasis on those groups in the lowest socioeconomic bracket.
Abstract: Each year, the American Cancer Society estimates the numbers of new cancer cases and deaths expected in the United States in the current year and compiles the most recent data on cancer incidence, mortality, and survival based on incidence data from the National Cancer Institute, the Centers for Disease Control and Prevention, and the North American Association of Central Cancer Registries and mortality data from the National Center for Health Statistics. A total of 1,638,910 new cancer cases and 577,190 deaths from cancer are projected to occur in the United States in 2012. During the most recent 5 years for which there are data (2004-2008), overall cancer incidence rates declined slightly in men (by 0.6% per year) and were stable in women, while cancer death rates decreased by 1.8% per year in men and by 1.6% per year in women. Over the past 10 years of available data (1999-2008), cancer death rates have declined by more than 1% per year in men and women of every racial/ethnic group with the exception of American Indians/Alaska Natives, among whom rates have remained stable. The most rapid declines in death rates occurred among African American and Hispanic men (2.4% and 2.3% per year, respectively). Death rates continue to decline for all 4 major cancer sites (lung, colorectum, breast, and prostate), with lung cancer accounting for almost 40% of the total decline in men and breast cancer accounting for 34% of the total decline in women. The reduction in overall cancer death rates since 1990 in men and 1991 in women translates to the avoidance of about 1,024,400 deaths from cancer. Further progress can be accelerated by applying existing cancer control knowledge across all segments of the population, with an emphasis on those groups in the lowest socioeconomic bracket.
10,630 citations
TL;DR: A standard set of quantity names and symbols related to the estimation of kinetic parameters from dynamic contrast‐enhanced T1‐weighted magnetic resonance imaging data, using diffusable agents such as gadopentetate dimeglumine (Gd‐DTPA), are described.
Abstract: We describe a standard set of quantity names and symbols related to the estimation of kinetic parameters from dynamic contrast-enhanced T(1)-weighted magnetic resonance imaging data, using diffusable agents such as gadopentetate dimeglumine (Gd-DTPA). These include a) the volume transfer constant K(trans) (min(-1)); b) the volume of extravascular extracellular space (EES) per unit volume of tissue v(e) (0 < v(e) < 1); and c) the flux rate constant between EES and plasma k(ep) (min(-1)). The rate constant is the ratio of the transfer constant to the EES (k(ep) = K(trans)/v(e)). Under flow-limited conditions K(trans) equals the blood plasma flow per unit volume of tissue; under permeability-limited conditions K(trans) equals the permeability surface area product per unit volume of tissue. We relate these quantities to previously published work from our groups; our future publications will refer to these standardized terms, and we propose that these be adopted as international standards.
3,078 citations
TL;DR: This article develops ROC concepts in an intuitive way by identifying the fundamental issues that motivate ROC analysis and practical techniques for ROC data collection and data analysis.
Abstract: If the performance of a diagnostic imaging system is to be evaluated objectively and meaningfully, one must compare radiologists' image-based diagnoses with actual states of disease and health in a way that distinguishes between the inherent diagnostic capacity of the radiologists' interpretations of the images, and any tendencies to "under-read" or "over-read". ROC methodology provides the only known basis for distinguishing between these two aspects of diagnostic performance. After identifying the fundamental issues that motivate ROC analysis, this article develops ROC concepts in an intuitive way. The requirements of a valid ROC study and practical techniques for ROC data collection and data analysis are sketched briefly. A survey of the radiologic literature indicates the broad variety of evaluation studies in which ROC analysis has been employed.
1,780 citations
TL;DR: The use of CAD in the interpretation of screening mammograms can increase the detection of early-stage malignancies without undue effect on the recall rate or positive predictive value for biopsy.
Abstract: PURPOSE: To prospectively assess the effect of computer-aided detection (CAD) on the interpretation of screening mammograms in a community breast center. MATERIALS AND METHODS: Over a 12-month period, 12,860 screening mammograms were interpreted with the assistance of a CAD system. Each mammogram was initially interpreted without the assistance of CAD, followed immediately by a reevaluation of areas marked by the CAD system. Data were recorded to measure the effect of CAD on the recall rate, positive predictive value for biopsy, cancer detection rate, and stage of malignancies at detection. RESULTS: When comparing the radiologist’s performance without CAD with that when CAD was used, the authors observed the following: (a) an increase in recall rate from 6.5% to 7.7%, (b) no change in the positive predictive value for biopsy at 38%, (c) a 19.5% increase in the number of cancers detected, and (d) an increase in the proportion of early-stage (0 and I) malignancies detected from 73% to 78%. CONCLUSION: The u...
801 citations