Utility of quantitative measurement ofT2using Restriction Spectrum Imaging for detection of clinically significant prostate cancer

Utility of quantitative measurement of T2 using Restriction Spectrum Imaging for detection of clinically significant prostate cancer | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search Utility of quantitative measurement of T 2 using Restriction Spectrum Imaging for detection of clinically significant prostate cancer Mariluz Rojo Domingo , Christopher C Conlin , Roshan A Karunamuni , Courtney Ollison , Madison T Baxter , Karoline Kallis , Deondre D Do , Yuze Song , Joshua M Kuperman , Ahmed S Shabaik , Michael E Hahn , Paul M Murphy , Rebecca Rakow-Penner , Anders M Dale , Tyler M Seibert doi: https://doi.org/10.1101/2024.03.29.24305033 Mariluz Rojo Domingo 1 Department of Bioengineering, UC San Diego Jacobs School of Engineering , La Jolla, CA, USA 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA MS Find this author on Google Scholar Find this author on PubMed Search for this author on this site Christopher C Conlin 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Roshan A Karunamuni 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Courtney Ollison 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA BS Find this author on Google Scholar Find this author on PubMed Search for this author on this site Madison T Baxter 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA MS Find this author on Google Scholar Find this author on PubMed Search for this author on this site Karoline Kallis 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Deondre D Do 1 Department of Bioengineering, UC San Diego Jacobs School of Engineering , La Jolla, CA, USA 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA BS Find this author on Google Scholar Find this author on PubMed Search for this author on this site Yuze Song 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA 4 Department of Electrical and Computer Engineering, UC San Diego Jacobs School of Engineering , La Jolla, CA, USA MS Find this author on Google Scholar Find this author on PubMed Search for this author on this site Joshua M Kuperman 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ahmed S Shabaik 5 Department of Pathology, UC San Diego School of Medicine , La Jolla, CA, USA MD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Michael E Hahn 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA MD ,PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Paul M Murphy 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Rebecca Rakow-Penner 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Anders M Dale 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA 6 Department of Neurosciences, UC San Diego School of Medicine , La Jolla, CA, USA 7 Halıcıoğlu Data Science Institute , UC San Diego, La Jolla, CA, USA PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site Tyler M Seibert 1 Department of Bioengineering, UC San Diego Jacobs School of Engineering , La Jolla, CA, USA 2 Department of Radiation Medicine and Applied Sciences, UC San Diego School of Medicine , La Jolla, CA, USA 3 Department of Radiology, UC San Diego School of Medicine , La Jolla, CA, USA MD, PhD Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: tseibert{at}health.ucsd.edu Abstract Full Text Info/History Metrics Data/Code Preview PDF Abstract Background The Restriction Spectrum Imaging restriction score (RSIrs) has demonstrated higher diagnostic accuracy for clinically significant prostate cancer (csPCa) than conventional DWI. Both diffusion and T 2 properties of prostate tissue inform the RSI signal, and studies have shown that each may be valuable for csPCa discrimination. Purpose To determine whether prostate T 2 varies across RSI compartments and in the presence of csPCa, and to evaluate whether consideration of compartmental T 2 (c T 2 ) improves csPCa detection over RSIrs alone. Study Type Retrospective. Population Two cohorts (46 and 195 patients) scanned for csPCa. Field Strength/Sequence 3T multi- b -value DWI acquired at multiple TEs. Assessment c T 2 values were computed from multi-TE RSI data and compared between RSI model compartments. csPCa detection was compared between RSIrs and a logistic regression model (LRM) for predicting the probability of csPCa using c T 2 in combination with RSI measurements. Statistical Tests T wo-sample t-tests (α=0.05) were used to compare c T 2 values between compartments and between patients with and without csPCa. Area under the receiver operating characteristic curve (AUC) was used to evaluate csPCa detection performance. Results In both cohorts, T 2 differed ( p <0.05) across all RSI compartments ( C 1, C 2, C 3, C 4 ). Voxel-level data from cohort 1 showed that T 2 differed between normal and cancerous tissue in C 1, C 2, C 3 ( p <0.001). Whole-prostate T 2 differed between patients with and without csPCa in C 3 ( p =0.02). In cohort 2, whole-prostate T 2 differed in C 1 ( p =0.01), C 3 ( p =0.01), and C 4 ( p <0.01). Consideration of c T 2 improved csPCa discrimination compared to diffusion alone, but not compared to RSIrs [cohort 1: 0.80 vs 0.70 (diffusion) and 0.80 (RSIrs), cohort 2: 0.72 vs 0.65 (diffusion) and 0.72 (RSIrs)]. Data Conclusion Significant differences in c T 2 were observed between normal and cancerous prostatic tissue. With our data, however, consideration of c T 2 did not significantly improve cancer detection performance over RSIrs alone. Introduction Multiparametric magnetic resonance imaging (mpMRI) has become an important tool for the diagnosis of prostate cancer (PCa) 1 . MpMRI has proven to reduce unnecessary biopsies, mitigates overdiagnosis of clinically insignificant prostate cancer (indolent PCa), and enhances detection of clinically significant prostate cancer (csPCa, Grade Group 2 or higher) 2 , 3 . In standard reporting of prostate MRI (Prostate Imaging Reporting and Data System, PI-RADS v2.1 4 ), diffusion-weighted imaging (DWI) and T 2 -weighted imaging are the principal modalities used to detect csPCa. DWI measures the random movement of water molecules within tissues, aiding in the visualization of areas of restricted diffusion associated with hypercellular csPCa 5 . Meanwhile, T 2 -weighted imaging provides detailed anatomical information and facilitating the visualization of abnormalities in prostate tissue. The combined analysis of these two sequences allows for a more in-depth evaluation of potential tumor lesions, thereby contributing to csPCa detection and characterization. The challenge with the interpretation of conventional mpMRI lies in its inherent subjectivity and variability 6 . The interpretations of imaging data by different radiologists that rely on qualitative assessment alone leads to inconsistencies in the identification and characterization of csPCa lesions 7 . Interobserver variability significantly limits the accuracy and reliability of csPCa diagnosis 8 . To enhance diagnostic accuracy, there is a growing emphasis on the development and adoption of quantitative MRI approaches. Quantitative MRI aims to provide objective metrics of tissue properties associated with the probability of csPCa, offering the potential for more standardized and reproducible image assessment. Restriction Spectrum Imaging (RSI) is a quantitative approach to DWI for csPCa detection and characterization. RSI scans are acquired at multiple b -values (diffusion weightings) to distinguish diffusion signal from tissue micro-compartments (intracellular water, extracellular hindered water, freely diffusing water, and flowing fluid) 9 – 12 . However, RSI models typically do not incorporate quantitative T 2 measurements. On the other hand, studies using luminal water imaging (LWI) and hybrid multidimensional MRI have shown that tissue T 2 can differ between prostate tissue compartments and provide diagnostic information that is complementary to diffusion 13 , 14 . In this study, we acquired prostate RSI data at multiple echo times (TEs) to measure compartmental T 2 in addition to diffusion. We aimed to determine whether compartmental T 2 (i.e., T 2 within each RSI micro-compartment) differs between cancerous and normal prostate tissue, and whether consideration of compartmental T 2 in RSI yields improved detection of csPCa. Methods Study Population This study was approved by the institutional review board (IRB). For the prospective cohort, a waiver of consent was obtained to acquire images with two TEs for image optimization, as well as to access patient images and other clinical information. Consecutive patients of age 18 years or older who underwent MRI with multi-TE RSI for suspected PCa were included in the two cohorts of this study. The first cohort (retrospective) included patients scanned for PCa between August and December of 2016 with multi-TE RSI as part of a quality improvement project. The second cohort included patients who were scanned for PCa between March 2021 and January of 2023. Additionally, the second cohort included participants in a separate, prospective study (also approved by the IRB) of patients with localized prostate cancer using the same multi-TE RSI protocol. These latter participants gave informed consent to undergo additional imaging before and after radiation therapy; only pre-treatment imaging was included in the present study. Patients were excluded if they had received any treatment for PCa prior to the MRI acquisition or if a lesion with PI-RADS score ≥3 was detected on MRI but no biopsy information was available. Routine Clinical Evaluation Most patients in both cohorts underwent prostate MRI as part of routine clinical care for PCa. MpMRI was performed according to PI-RADS guidelines, and interpretation was made per clinical routine using PI-RADS v2.1. Several patients, all from cohort 2, had PCa diagnosed on systematic biopsy without MRI, and then had an MRI with RSI before any treatment as part of a prospective study; PI-RADS scores are not available for these subjects. The presence of clinically significant prostate cancer (csPCa, grade group 2 or higher) was determined from the best available histopathology, either biopsy (typically systematic 12-core biopsy with additional targeted cores for suspicious lesions on MRI) or radical prostatectomy. Patients with PI-RADS lesions of 1 or 2 with no biopsy were considered negative for csPCa, in accordance with European Association of Urology (EAU) guidelines 15 – 17 . RSI Data Acquisition and Processing All patients were scanned with an expanded MRI protocol that included two multi- b -value RSI acquisitions performed with different TEs. MRI acquisition details are summarized in Table 1 . All MR imaging was performed on a 3T clinical scanner (Discovery MR750; GE Healthcare, Waukesha, WI, USA), using a 32-channel phased-array coil over the pelvis. For each patient, two axial, multi- b -value DWI volumes were separately acquired using two different TEs: 80 ms and 100 ms for cohort 1, and 76 ms and 90 ms for cohort 2. The other parameters were the same between scans. In addition to the DWI volumes, a single T 2 -weighted volume was acquired for anatomical reference using the same scan coverage as the DWI volumes. MRI post-processing was performed using programs implemented in MATLAB (MathWorks, Natick, MA, USA). DWI volumes were corrected to account for B 0 -inhomogeneities, gradient nonlinearities, eddy currents 18 , and image noise 10 . Samples at each b -value were averaged together. Image registration 19 was applied to correct for patient motion between acquisitions. View this table: View inline View popup Download powerpoint Table 1: Acquisition details for DWI and T 2 -weighted image volumes. All MR imaging was performed on a 3T clinical scanner (Discovery MR750; GE Healthcare), using a 32-channel phased-array coil over the pelvis. The two DWI volumes for both cohorts were acquired using different TEs to allow for examination of T 2 relaxation in prostatic tissue compartments. The single T 2 -weighted volume was acquired for anatomical reference. For patients in cohort 1, regions of interest (ROIs) were manually defined on T 2 -weighted images over the whole prostate, peripheral zone, and transition zone (the central zone was included with the transition zone). The contouring of the prostate zones and tumor lesions was performed using MIM software (MIM Software, Inc; Cleveland, OH, USA), by a radiation oncologist with 3 years of experience and two board-certified sub-specialist radiologists with 4 and 6 years of experience, using all available clinical imaging and pathologic information 11 . Radiologist-certified contours of the prostate zones and lesions were not obtainable for cohort 2. Instead, contours of the whole prostate were generated automatically using OnQ Prostate software from Cortechs Labs (Cortech.ai; San Diego, CA, USA). RSI modeling Prior studies established and validated a four-compartment RSI model of the diffusion signal10,11,20: S(b) denotes the measured DWI signal intensity at a particular b -value, which is modeled as a linear combination of exponential decays representing four diffusion compartments. C i describes the compartmental signal contributions to be determined via model-fitting. The diffusion coefficients, D i , are fixed for each of the four tissue compartments to empirically determined 10 values that broadly represent restricted, hindered, free diffusion, and vascular flow: 1.1e-4, 1.8e-3, 3.6e-3, and 0.1220 mm 2 /s, respectively. Signal-contribution ( C i ) maps were computed for both DWI volumes per patient by fitting this model to the signal-vs.- b -value curve from each voxel. A previously validated biomarker for PCa called the RSI restriction score (RSIrs) was computed by dividing the signal intensity of the restricted diffusion compartment, C 1 , at each voxel by the median signal intensity within the whole prostate on the b = 0 mm 2 /s DWI images (an index of apparent T 2 -weighting in the prostate) 11 , 20 – 23 . Compartmental T 2 mapping and analysis by csPCa status T 2 maps were computed for each compartment of the RSI model by fitting the monoexponential T 2 decay formula to the signal values from the two C i maps with different TEs. The median T 2 within the whole prostate was then computed. Two-sample t -tests ( α = 0.05) were used to determine whether there were significant differences in median T 2 between compartments. For cohort 1, two-sample t -tests ( α = 0.05) were used to compare median compartmental T 2 values between benign or clinically insignificant PCa tissue and csPCa lesions. For both cohorts, we used two-sample t -tests ( α = 0.05) to compare each patient’s median T 2 by compartment in the whole prostate, and whether compartmental T 2 was significantly different between patients with and without csPCa. Any compartments with a significant difference in T 2 between normal and cancerous tissue were noted for inclusion in subsequent multivariable modeling. Logistic regression model fitting and evaluation of cancer-detection performance A logistic regression model (LRM) was developed to estimate the probability that a given voxel of tissue contains csPCa given measurements of diffusion and compartmental T 2 . RSIrs 11 was included as the diffusion parameter of the model. Compartmental T 2 was included in the LRM for each compartment that showed a significant difference in T 2 between normal and cancerous tissue. Cohort 1 had radiologist-certified lesion contours available and was therefore used to train the LRM. In patients with csPCa, voxels inside the lesion contours were labeled as csPCa-positive, while prostate voxels outside the lesion contours were labeled as csPCa-negative. In these patients, voxels labeled as csPCa-positive were included to train the LRM and all non-csPCa voxels were excluded. In patients without csPCa, diffusion and compartmental T 2 measurements from all voxels within the entire prostate were used to train the LRM and labeled as csPCa-negative. Ten-fold cross-validation was performed to evaluate voxel-level csPCa-detection performance of the model within cohort 1. We assessed csPCa-detection performance using the area under the receiver operating characteristic curve (AUC) and calculated 95% confidence intervals (CI) from 10,000 bootstrap samples. Both cohorts were used to test the patient-level csPCa-detection performance of the model. For patient-level analysis of the LRM, the highest probability value observed within the whole prostate was used as the predictor variable. Similarly, the maximum RSIrs value within the whole prostate was used as the patient-level predictor for RSIrs. We also computed maximum C 1 to obtain the patient-level performance of diffusion only, as RSIrs incorporates global prostate T 2 signal in addition to diffusion signal. AUC values were computed for maximum C 1 , maximum RSIrs, and the LRM, and compared using two-sample t -tests ( α = 0.05). The 95% confidence intervals were estimated through random sampling with replacement from 10,000 bootstrap patient samples. Results Study Population Cohort 1 comprised 46 patients (age: 70 ± 10 years; PSA: 10.6 ±16.9 ng/mL). Cohort 2 comprised 195 patients (age: 69 ± 8 years; PSA: 8.2 ± 8.5 ng/mL). In cohort 1, 22 of 46 patients (47.8%) had csPCa, while the remaining 24 had either low-grade (grade group 1) disease or no cancer. In cohort 2, 96 of 195 (49.2%) patients had csPCa. 38 participants from cohort 2 had no PI-RADS scores available because csPCa was diagnosed on systematic biopsy without MRI and then had an MRI with RSI acquisition as part of a separate, prospective study. Table 2 summarizes the patient characteristics of both cohorts. View this table: View inline View popup Download powerpoint Table 2. Summary of radiologic and pathologic characteristics of the two cohorts of patients included in this study. Compartmental T 2 mapping Figure 1 shows compartmental T 2 maps for two patients with csPCa, one from each cohort. Download figure Open in new tab Figure 1: RSI signal contribution ( C i ) and compartmental T 2 (c T 2 ) maps for two patients with csPCa. The top panel corresponds to a patient from cohort 1 and the bottom panel to a patient from cohort 2. Compartmental T 2 maps were computed for each RSI model compartment by measuring the T 2 -weighted signal decay of the signal-contribution map for different TEs. The signal-contribution maps shown here were computed from the DWI acquisition with shorter TE. Whole-mount (WM) histopathology results were available for the patient from cohort 2 and illustrate the lesion contour in the prostate. Figure 2 shows violin plots of median T 2 within the whole prostate for each RSI model compartment. In both cohorts, the highest median T 2 values were observed in C 3 , followed by C 2 , C 4 , and finally C 1 . Compartmental T 2 was significantly different between any two compartments ( p < 0.05). Download figure Open in new tab Figure 2: Violin plots showing the distribution of median T 2 values in the whole prostate for each of the four RSI model compartments. Within each cohort, compartmental T 2 was significantly different between any two compartments ( p < 0.05). Left panel: cohort 1 (n = 46). Right panel: cohort 2 (n = 195). Voxel-level analysis of compartmental T 2 values by csPCa status For each compartment of the RSI model, the comparison of T 2 values between csPCa and prostate tissue outside of csPCa lesions is illustrated in Figure 3 . This figure corresponds to cohort 1, which has radiologist-certified csPCa lesion contours. csPCa lesions showed significantly higher compartmental T 2 values in compartment 1 ( p -value << 0.001) and compartment 2 ( p -value << 0.001) than normal tissue. In addition, the compartmental T 2 values of compartment C 3 were significantly lower in csPCa lesions ( p -value << 0.001). The compartmental T 2 values for compartment C 4 were not significantly different between csPCa and normal tissues ( p -value = 0.17). Download figure Open in new tab Figure 3: Violin plots showing the distribution of median compartmental T 2 values from cohort 1 in both csPCa lesions and the surrounding prostate tissue. The red “+” corresponds to the mean value in each violin. Each panel corresponds to one of the RSI diffusion compartments. csPCa lesions showed significantly higher compartmental T2 values in C 1 ( p -value << 0.001) and C 2 ( p -value << 0.001) than tissues outside of csPCa lesions. Compartmental T 2 values of C 3 were also significantly lower in csPCa lesions than outside csPCa lesions ( p -value << 0.001). A red star indicates a significant difference ( p- value < 0.05) in compartmental T 2 between csPCa lesions and the prostate tissue outside the lesions. WP: whole prostate. Patient-level analysis of compartmental T 2 values by csPCa status The comparison of compartmental T 2 values between patients with and without csPCa is shown in Figure 4 . In both cohorts, patients with csPCa had higher compartmental T 2 values in compartment C 1 than patients with no csPCa. In cohort 2, median C 1 compartmental T 2 was significantly higher ( p -value = 0.07 for cohort 1; p -value = 0.01 for cohort 2). Median C 3 compartmental T 2 was significantly different between csPCa and patients without csPCa in both cohorts ( p -value = 0.02 for cohort 1; p -value = 0.01 for cohort 2). Median C 4 compartmental T 2 was also significantly different in cohort 2 ( p -value << 0.01). Compartmental T 2 values for the other compartments were not significantly different between patients with and without csPCa. Download figure Open in new tab Figure 4: Violin plots comparing compartmental T 2 within the whole prostate between patients with csPCa and those without. A red star indicates a significant difference ( p- value < 0.05) in whole-prostate compartmental T 2 between csPCa and non-csPCa patients. Left panel: cohort 1 (n = 46). Right panel: cohort 2 (n = 195). Logistic regression model fitting and evaluation of csPCa-detection performance T 2 measurements from RSI compartments 1, 2, and 3 ( C 1 , C 2 , C 3 ) were included as the T 2 parameters of the LRM. These three compartments showed significantly different median T 2 signal between in csPCa lesions vs csPCa-negative voxels. The LRM predictors were RSIrs, C 1 T 2 , C 2 T 2 , and C 3 T 2 . The model coefficients with 95% confidence interval for the y-intercept and RSIrs were 6.367 (6.323, 6.416) and -51.621 (-52.354, -50.888), respectively. The weights for C 1 T 2 , C 2 T 2 , and C 3 T 2 were < 0.005. Example probability maps computed from the model are shown in Figure 5 for two patients with csPCa, alongside maps of RSIrs. Download figure Open in new tab Figure 5: RSI restriction score and multivariable model probability maps of the prostate for patients with csPCa. Top panel: Patient from cohort 1 with a lesion in the left peripheral zone. Bottom panel: Patient from cohort 2 with a lesion in the transition zone. The multivariable model uses compartmental T 2 measurements from each voxel in addition to the RSI restriction score to determine the probability that it contains csPCa. Voxel-level cancer detection For voxel-level cancer detection, the 10-fold cross-validation mean AUC of the LRM was 0.98 [95% CI: 0.957, 0.985], versus 0.98 [95% CI: 0.958, 0.986] for maximum RSIrs, indicating that incorporating compartmental T 2 did not improve discrimination over RSIrs alone ( p -value = 0.9). Patient-level cancer detection For cohort 1, the AUC of the LRM was 0.804 [0.648,0.930], versus 0.805 [0.648, 0.931] for RSIrs. The mean AUC for maximum C 1 was 0.695 [0.530, 0.851]. The difference in AUCs between RSIrs and the multivariable model was not significantly different ( p -value = 0.26). Both RSIrs and the LRM performed significantly better than diffusion (maximum C 1 ) alone (both p -values << 0.001). For cohort 2, the mean LRM AUC for 10,000 bootstrapped samples was 0.724 [0.650, 0.793]. The mean AUC of RSIrs for 10,000 bootstrapped samples was 0.725 [0.652, 0.794]. For maximum C 1 the mean AUC was 0.654 [0.573,0.730]. The difference in AUCs between RSIrs and the multivariable model was not significantly different ( p -value = 0.54). Both RSIrs and the LRM performed significantly better than diffusion (maximum C 1 ) alone (both p -values << 0.001). Download figure Open in new tab Figure 5: Patient-level diagnostic performance of maximum RSIrs, the LRM and maximum C 1 (max C 1 ) for cohorts 1 and 2. For cohort 1, mean AUC values were 0.695 for maximum C 1 , 0.805 for RSIrs and 0.804 for the multivariable model. For cohort 2, mean AUC values were 0.654 for maximum C 1 , 0.725 for RSIrs and 0.724 for the multivariable model. Discussion We found that compartmental T 2 values were significantly different across RSI diffusion compartments. Moreover, compartmental T 2 values differed between csPCa lesions and benign tissue or low-grade PCa in compartments C 1, C 2, and C 3 . At the patient level, there were also differences in whole-prostate T 2 between patients with no csPCa and those with biopsy-proven csPCa. Quantitative differences in compartmental T 2 may provide insight into the microstructural changes associated with PCa. For example, extracellular matrix remodeling may contribute to the increased compartmental T 2 observed in C 2 12 . Lower compartmental T 2 values in C 3 may reflect hyperplasia-induced reductions in luminal space. 13 While the elevated compartmental T 2 observed in C 1 of csPCa patients may seem opposed to the known overall hypointense appearance of csPCa on T 2 -weighted MRI, it is consistent with an increase in nuclear volume fraction that is typical of cancer cells 24 . We have previously shown that RSIrs is a useful quantitative DWI biomarker for csPCa at the voxel and patient levels 11 , 20 , 22 . Here, we evaluated whether incorporating compartmental T 2 values would improve csPCa discrimination. We demonstrated that T 2 effects have csPCa discriminatory value, as they showed higher AUC values compared to diffusion alone in both cohorts. However, consideration of compartmental T 2 did not significantly improve csPCa-detection performance over maximum RSIrs at the voxel- or patient-level. RSIrs already includes information related to global T 2 signal in the prostate (median signal in the whole prostate on the b =0 acquisition). It appears that this global T 2 assessment may contain sufficient information to normalize RSI diffusion signal and improve csPCa discrimination. This result suggests that compartmental T 2 mapping may not be worth the additional scan time and computation required. Other advanced imaging techniques have also been used to measure T 2 in the prostate. LWI utilizes the unique T 2 relaxation rates associated with various components of prostate tissue to quantify the fractional volume of glandular lumen, denoted as luminal water fraction (LWF) 25 . This method takes advantage of the observed alterations in the composition of prostatic tissue in the presence of cancer and the Gleason grade of the cancer for PCa diagnosis 26 . Hybrid multidimensional MRI (HM-MRI) exploits the interdependence of T 2 and ADC values to measure prostate volume fractions of the lumen, epithelium, and stroma 27 . Studies using LWI and HM-MRI demonstrate lower T 2 values in cancer lesions compared to normal prostate tissue with increasing Gleason Grade. The decrease in T 2 results from a decrease in luminal volume due to cellular hyperplasia 26 , 28 . This trend agrees with the decrease observed in this study of T 2 in compartment C 3 of patients with csPCa. This compartment reflects signal from freely diffusing water in the prostate, which we expect to find predominantly in luminal tissue and to be impacted by a reduction in luminal space. Prior work with HM-MRI showed that the decrease in luminal space is largely the result of epithelial tissue proliferation, indicated by an increase in the measured epithelial volume fraction 27 . Since RSI does not explicitly assign signal contributions to a particular tissue type, this aspect of HM-MRI is harder to align with the present study. However, we can be sure that the changes observed in the T 2 of compartment C 1 reflect, at least in part, an increase in the overall tissue cellularity 29 . Signal contributions in this compartment are also dependent upon the nuclear volume fraction of cells in the tissue 24 , and the increase in the T 2 observed in C 1 of csPCa patients suggests an increase in nuclear volume fraction. Whole-mount histopathology data is currently being collected as part of an ongoing study to map changes in compartmental RSI signal to the histological restructuring of prostate tissue due to csPCa. Future work will also evaluate larger patient cohorts to consider the potential impact of compartmental T 2 measurements on the prediction of csPCa grade and tumor extent with the RSI framework. Finally, an ongoing, prospective trial is evaluating the impact of RSIrs and compartmental T 2 on csPCa diagnosis in multiple centers and with multiple readers. Limitations The use of separate acquisitions and only two TEs may have limited the accuracy of voxel-wise compartmental T 2 measurements. Our study used two TEs for each cohort, where the lowest TE was 76 ms. Other quantitative MRI approaches use lower TEs (less than 30 ms) for T 2 mapping 14 , 26 . The high b -values necessary to optimally estimate RSI C 1 are generally incompatible with very low TE on clinical scanners, so separate acquisitions would be required to evaluate the combination of optimal RSIrs and T 2 estimated with very low TE. Another limitation of this study is that our voxel-level data included only high-confidence csPCa and control categories, leaving little room for improvement over RSIrs for the voxel-level analysis 11 . Conclusion T 2 mapping affords insights into characteristics of benign and cancerous prostate tissue, but we did not find compelling evidence that acquisitions with multiple TE is necessary for patient-level csPCa detection with RSI. Data Availability All data produced in the present study are available upon reasonable request to the authors. Footnotes The only thing I added is the "MS" after Mariluz Rojo Domingo in the author list. References 1. ↵ Leung DKW , Chiu PKF , Ng CF , Teoh JYC . Role of pre-biopsy multiparametric MRI in prostate cancer diagnosis: Evidence from the literature . Turk J Urol . 2021 ; 47 ( Suppl 1 ): S65 – S70 . doi: 10.5152/tud.2020.20360 OpenUrl CrossRef 2. ↵ de Rooij M , Hamoen EHJ , Fütterer JJ , Barentsz JO , Rovers MM . Accuracy of multiparametric MRI for prostate cancer detection: a meta-analysis . AJR Am J Roentgenol . 2014 ; 202 ( 2 ): 343 – 351 . doi: 10.2214/AJR.13.11046 OpenUrl CrossRef PubMed 3. ↵ Schoots IG , Roobol MJ , Nieboer D , Bangma CH , Steyerberg EW , Hunink MGM . Magnetic resonance imaging-targeted biopsy may enhance the diagnostic accuracy of significant prostate cancer detection compared to standard transrectal ultrasound-guided biopsy: a systematic review and meta-analysis . Eur Urol . 2015 ; 68 ( 3 ): 438 – 450 . doi: 10.1016/j.eururo.2014.11.037 OpenUrl CrossRef PubMed 4. ↵ Scott R , Misser SK , Cioni D , Neri E . PI-RADS v2.1: What has changed and how to report . SA J Radiol . 2021 ; 25 ( 1 ): 2062 . doi: 10.4102/sajr.v25i1.2062 OpenUrl CrossRef 5. ↵ Demirel HC , Davis JW . Multiparametric magnetic resonance imaging: Overview of the technique, clinical applications in prostate biopsy and future directions . Turk J Urol . 2018 ; 44 ( 2 ): 93 – 102 . doi: 10.5152/tud.2018.56056 OpenUrl CrossRef 6. ↵ Pickersgill NA , Vetter JM , Andriole GL , et al. Accuracy and Variability of Prostate Multiparametric Magnetic Resonance Imaging Interpretation Using the Prostate Imaging Reporting and Data System: A Blinded Comparison of Radiologists . Eur Urol Focus . 2020 ; 6 ( 2 ): 267 – 272 . doi: 10.1016/j.euf.2018.10.008 OpenUrl CrossRef 7. ↵ Midiri F , Vernuccio F , Purpura P , Alongi P , Bartolotta TV . Multiparametric MRI and Radiomics in Prostate Cancer: A Review of the Current Literature . Diagn Basel Switz . 2021 ; 11 ( 10 ): 1829 . doi: 10.3390/diagnostics11101829 OpenUrl CrossRef 8. ↵ Stabile A , Giganti F , Kasivisvanathan V , et al. Factors Influencing Variability in the Performance of Multiparametric Magnetic Resonance Imaging in Detecting Clinically Significant Prostate Cancer: A Systematic Literature Review . Eur Urol Oncol . 2020 ; 3 ( 2 ): 145 . doi: 10.1016/j.euo.2020.02.005 OpenUrl CrossRef 9. ↵ Brunsing RL , Schenker-Ahmed NM , White NS , et al. Restriction spectrum imaging: An evolving imaging biomarker in prostate MRI: Prostate MRI with Restriction Spectrum Imaging: A Review . J Magn Reson Imaging . 2017 ; 45 ( 2 ): 323 – 336 . doi: 10.1002/jmri.25419 OpenUrl CrossRef 10. ↵ Conlin CC , Feng CH , RodrÃ-guez-Soto AE , et al. Improved Characterization of Diffusion in Normal and Cancerous Prostate Tissue through Optimization of the Restriction Spectrum Imaging Signal Model . Radiology and Imaging ; 2020 . doi: 10.1101/2020.03.27.20042069 OpenUrl Abstract / FREE Full Text 11. ↵ Feng CH , Conlin CC , Batra K , et al. Voxel-level Classification of Prostate Cancer on Magnetic Resonance Imaging: Improving Accuracy Using FOUR-COMPARTMENT Restriction Spectrum Imaging . J Magn Reson Imaging . 2021 ; 54 ( 3 ): 975 – 984 . doi: 10.1002/jmri.27623 OpenUrl CrossRef 12. ↵ White NS , McDonald CR , Farid N , et al. Diffusion-Weighted Imaging in Cancer: Physical Foundations and Applications of Restriction Spectrum Imaging . Cancer Res . 2014 ; 74 ( 17 ): 4638 – 4652 . doi: 10.1158/0008-5472.CAN-13-3534 OpenUrl Abstract / FREE Full Text 13. ↵ Hectors SJ , Said D , Gnerre J , Tewari A , Taouli B . Luminal Water Imaging: Comparison With Diffusion-Weighted Imaging (DWI) and PI-RADS for Characterization of Prostate Cancer Aggressiveness . J Magn Reson Imaging . 2020 ; 52 ( 1 ): 271 – 279 . doi: 10.1002/jmri.27050 OpenUrl CrossRef 14. ↵ Chatterjee A , Bourne RM , Wang S , et al. Diagnosis of Prostate Cancer with Noninvasive Estimation of Prostate Tissue Composition by Using Hybrid Multidimensional MR Imaging: A Feasibility Study . Radiology . 2018 ; 287 ( 3 ): 864 – 873 . doi: 10.1148/radiol.2018171130 OpenUrl CrossRef 15. ↵ EAU Guidelines on Prostate Cancer - Uroweb . Uroweb - European Association of Urology . Accessed October 16, 2023. https://uroweb.org/guidelines/prostate-cancer 16. Abdul Raheem R , Razzaq A , Beraud V , et al. Can a prostate biopsy be safely deferred on PI-RADS 1,2 or 3 lesions seen on pre-biopsy mp-MRI? Arab J Urol . 21 ( 1 ): 10 – 17 . doi: 10.1080/2090598X.2022.2119711 OpenUrl CrossRef 17. ↵ Reijnen JS , Marthinsen JB , Tysland AO , et al. Results from a PI-RADS-based MRI-directed diagnostic pathway for biopsy-naive patients in a non-university hospital . Abdom Radiol N Y . 2021 ; 46 ( 12 ): 5639 – 5646 . doi: 10.1007/s00261-021-03249-8 OpenUrl CrossRef 18. ↵ Holland D , Kuperman JM , Dale AM . Efficient correction of inhomogeneous static magnetic field-induced distortion in Echo Planar Imaging . NeuroImage . 2010 ; 50 ( 1 ): 175 – 183 . doi: 10.1016/j.neuroimage.2009.11.044 OpenUrl CrossRef PubMed 19. ↵ Paquin D , Levy D , Schreibmann E , Xing L . MULTISCALE IMAGE REGISTRATION . 20. ↵ Zhong AY , Digma LA , Hussain T , et al. Automated Patient-level Prostate Cancer Detection with Quantitative Diffusion Magnetic Resonance Imaging . Eur Urol Open Sci . 2023 ; 47 : 20 – 28 . doi: 10.1016/j.euros.2022.11.009 OpenUrl CrossRef 21. Conlin CC , Karunamuni R , Hussain TS , et al. Background Prostate Tissue Is Quantitatively Abnormal on MRI in Patients with Clinically Significant Prostate Cancer . Radiology and Imaging ; 2022 . doi: 10.1101/2022.10.12.22280855 OpenUrl Abstract / FREE Full Text 22. ↵ Lui AJ , Kallis K , Zhong AY , et al. ReIGNITE RT Boost: An International Study Testing the Accuracy and Feasibility of Using Restriction Spectrum Imaging (RSI) MRI to Guide Intraprostatic Tumor Target Volume for Radiotherapy Boost . Oncology ; 2022 . doi: 10.1101/2022.12.13.22283420 OpenUrl Abstract / FREE Full Text 23. ↵ Kallis K , Conlin CC , Zhong AY , et al. Comparison of Synthesized and Acquired High b -Value Diffusion-Weighted MRI for Detection of Prostate Cancer . Radiology and Imaging ; 2023 . doi: 10.1101/2023.02.17.23286100 OpenUrl Abstract / FREE Full Text 24. ↵ White NS , Dale AM . Distinct effects of nuclear volume fraction and cell diameter on high b-value diffusion MRI contrast in tumors: Diffusion in Tumor Cells . Magn Reson Med . 2014 ; 72 ( 5 ): 1435 – 1443 . doi: 10.1002/mrm.25039 OpenUrl CrossRef PubMed 25. ↵ Chatterjee A , Harmath C , Oto A . New prostate MRI techniques and sequences . Abdom Radiol . 2020 ; 45 ( 12 ): 4052 – 4062 . doi: 10.1007/s00261-020-02504-8 OpenUrl CrossRef PubMed 26. ↵ Sabouri S , Chang SD , Savdie R , et al. Luminal Water Imaging: A New MR Imaging T2 Mapping Technique for Prostate Cancer Diagnosis . Radiology . 2017 ; 284 ( 2 ): 451 – 459 . doi: 10.1148/radiol.2017161687 OpenUrl CrossRef 27. ↵ Lee GH , Chatterjee A , Karademir I , et al. Comparing Radiologist Performance in Diagnosing Clinically Significant Prostate Cancer with Multiparametric versus Hybrid Multidimensional MRI . Radiology . 2022 ; 305 ( 2 ): 399 – 407 . doi: 10.1148/radiol.211895 OpenUrl CrossRef 28. ↵ Chatterjee A , Watson G , Myint E , Sved P , McEntee M , Bourne R . Changes in Epithelium, Stroma, and Lumen Space Correlate More Strongly with Gleason Pattern and Are Stronger Predictors of Prostate ADC Changes than Cellularity Metrics . Radiology . 2015 ; 277 ( 3 ): 751 – 762 . doi: 10.1148/radiol.2015142414 OpenUrl CrossRef 29. ↵ Liss MA , White NS , Parsons JK , et al. MRI-Derived Restriction Spectrum Imaging Cellularity Index is Associated with High Grade Prostate Cancer on Radical Prostatectomy Specimens . Front Oncol . 2015 ; 5 . doi: 10.3389/fonc.2015.00030 OpenUrl CrossRef View the discussion thread. Back to top Previous Next Posted April 09, 2024. Download PDF Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Utility of quantitative measurement of T2 using Restriction Spectrum Imaging for detection of clinically significant prostate cancer Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Utility of quantitative measurement of T 2 using Restriction Spectrum Imaging for detection of clinically significant prostate cancer Mariluz Rojo Domingo , Christopher C Conlin , Roshan A Karunamuni , Courtney Ollison , Madison T Baxter , Karoline Kallis , Deondre D Do , Yuze Song , Joshua M Kuperman , Ahmed S Shabaik , Michael E Hahn , Paul M Murphy , Rebecca Rakow-Penner , Anders M Dale , Tyler M Seibert medRxiv 2024.03.29.24305033; doi: https://doi.org/10.1101/2024.03.29.24305033 Share This Article: Copy Citation Tools Utility of quantitative measurement of T 2 using Restriction Spectrum Imaging for detection of clinically significant prostate cancer Mariluz Rojo Domingo , Christopher C Conlin , Roshan A Karunamuni , Courtney Ollison , Madison T Baxter , Karoline Kallis , Deondre D Do , Yuze Song , Joshua M Kuperman , Ahmed S Shabaik , Michael E Hahn , Paul M Murphy , Rebecca Rakow-Penner , Anders M Dale , Tyler M Seibert medRxiv 2024.03.29.24305033; doi: https://doi.org/10.1101/2024.03.29.24305033 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Radiology and Imaging Subject Areas All Articles Addiction Medicine (573) Allergy and Immunology (865) Anesthesia (302) Cardiovascular Medicine (4453) Dentistry and Oral Medicine (444) Dermatology (383) Emergency Medicine (609) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1515) Epidemiology (15242) Forensic Medicine (30) Gastroenterology (1131) Genetic and Genomic Medicine (6615) Geriatric Medicine (669) Health Economics (1001) Health Informatics (4552) Health Policy (1372) Health Systems and Quality Improvement (1614) Hematology (543) HIV/AIDS (1270) Infectious Diseases (except HIV/AIDS) (15929) Intensive Care and Critical Care Medicine (1106) Medical Education (624) Medical Ethics (147) Nephrology (670) Neurology (6625) Nursing (346) Nutrition (999) Obstetrics and Gynecology (1148) Occupational and Environmental Health (957) Oncology (3344) Ophthalmology (979) Orthopedics (369) Otolaryngology (421) Pain Medicine (436) Palliative Medicine (130) Pathology (665) Pediatrics (1696) Pharmacology and Therapeutics (693) Primary Care Research (714) Psychiatry and Clinical Psychology (5461) Public and Global Health (9252) Radiology and Imaging (2207) Rehabilitation Medicine and Physical Therapy (1371) Respiratory Medicine (1197) Rheumatology (597) Sexual and Reproductive Health (715) Sports Medicine (530) Surgery (714) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a02cf4175ee20f09',t:'MTc3OTk2ODEwMQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();