Deep learning-based image quality assessment: impact on detection accuracy of prostate cancer extraprostatic extension on MRI.

OBJECTIVE: To assess impact of image quality on prostate cancer extraprostatic extension (EPE) detection on MRI using a deep learning-based AI algorithm. MATERIALS AND METHODS: This retrospective, single institution study included patients who were imaged with mpMRI and subsequently underwent radical prostatectomy from June 2007 to August 2022. One genitourinary radiologist prospectively evaluated each patient using the NCI EPE grading system. Each T2WI was classified as low- or high-quality by a previously developed AI algorithm. Fisher's exact tests were performed to compare EPE detection metrics between low- and high-quality images. Univariable and multivariable analyses were conducted to assess the predictive value of image quality for pathological EPE. RESULTS: A total of 773 consecutive patients (median age 61 [IQR 56-67] years) were evaluated. At radical prostatectomy, 23% (180/773) of patients had EPE at pathology, and 41% (131/318) of positive EPE calls on mpMRI were confirmed to have EPE. The AI algorithm classified 36% (280/773) of T2WIs as low-quality and 64% (493/773) as high-quality. For EPE grade ≥ 1, high-quality T2WI significantly improved specificity for EPE detection (72% [95% CI 67-76%] vs. 63% [95% CI 56-69%], P = 0.03), but did not significantly affect sensitivity (72% [95% CI 62-80%] vs. 75% [95% CI 63-85%]), positive predictive value (44% [95% CI 39-49%] vs. 38% [95% CI 32-43%]), or negative predictive value (89% [95% CI 86-92%] vs. 89% [95% CI 85-93%]). Sensitivity, specificity, PPV, and NPV for EPE grades ≥ 2 and ≥ 3 did not show significant differences attributable to imaging quality. For NCI EPE grade 1, high-quality images (OR 3.05, 95% CI 1.54-5.86; P < 0.001) demonstrated a stronger association with pathologic EPE than low-quality images (OR 1.76, 95% CI 0.63-4.24; P = 0.24). CONCLUSION: Our study successfully employed a deep learning-based AI algorithm to classify image quality of prostate MRI and demonstrated that better quality T2WI was associated with more accurate prediction of EPE at final pathology.

A selective CutMix approach improves generalizability of deep learning-based grading and risk assessment of prostate cancer.

The Gleason score is an important predictor of prognosis in prostate cancer. However, its subjective nature can result in over- or under-grading. Our objective was to train an artificial intelligence (AI)-based algorithm to grade prostate cancer in specimens from patients who underwent radical prostatectomy (RP) and to assess the correlation of AI-estimated proportions of different Gleason patterns with biochemical recurrence-free survival (RFS), metastasis-free survival (MFS), and overall survival (OS). Training and validation of algorithms for cancer detection and grading were completed with three large datasets containing a total of 580 whole-mount prostate slides from 191 RP patients at two centers and 6218 annotated needle biopsy slides from the publicly available Prostate Cancer Grading Assessment dataset. A cancer detection model was trained using MobileNetV3 on 0.5 mm × 0.5 mm cancer areas (tiles) captured at 10× magnification. For cancer grading, a Gleason pattern detector was trained on tiles using a ResNet50 convolutional neural network and a selective CutMix training strategy involving a mixture of real and artificial examples. This strategy resulted in improved model generalizability in the test set compared with three different control experiments when evaluated on both needle biopsy slides and whole-mount prostate slides from different centers. In an additional test cohort of RP patients who were clinically followed over 30 years, quantitative Gleason pattern AI estimates achieved concordance indexes of 0.69, 0.72, and 0.64 for predicting RFS, MFS, and OS times, outperforming the control experiments and International Society of Urological Pathology system (ISUP) grading by pathologists. Finally, unsupervised clustering of test RP patient specimens into low-, medium-, and high-risk groups based on AI-estimated proportions of each Gleason pattern resulted in significantly improved RFS and MFS stratification compared with ISUP grading. In summary, deep learning-based quantitative Gleason scoring using a selective CutMix training strategy may improve prognostication after prostate cancer surgery.

Deep learning-based whole-body PSMA PET/CT attenuation correction utilizing Pix-2-Pix GAN.

PURPOSE: Sequential PET/CT studies oncology patients can undergo during their treatment follow-up course is limited by radiation dosage. We propose an artificial intelligence (AI) tool to produce attenuation-corrected PET (AC-PET) images from non-attenuation-corrected PET (NAC-PET) images to reduce need for low-dose CT scans. METHODS: A deep learning algorithm based on 2D Pix-2-Pix generative adversarial network (GAN) architecture was developed from paired AC-PET and NAC-PET images. 18F-DCFPyL PSMA PET-CT studies from 302 prostate cancer patients, split into training, validation, and testing cohorts (n = 183, 60, 59, respectively). Models were trained with two normalization strategies: Standard Uptake Value (SUV)-based and SUV-Nyul-based. Scan-level performance was evaluated by normalized mean square error (NMSE), mean absolute error (MAE), structural similarity index (SSIM), and peak signal-to-noise ratio (PSNR). Lesion-level analysis was performed in regions-of-interest prospectively from nuclear medicine physicians. SUV metrics were evaluated using intraclass correlation coefficient (ICC), repeatability coefficient (RC), and linear mixed-effects modeling. RESULTS: Median NMSE, MAE, SSIM, and PSNR were 13.26%, 3.59%, 0.891, and 26.82, respectively, in the independent test cohort. ICC for SUVmax and SUVmean were 0.88 and 0.89, which indicated a high correlation between original and AI-generated quantitative imaging markers. Lesion location, density (Hounsfield units), and lesion uptake were all shown to impact relative error in generated SUV metrics (all p < 0.05). CONCLUSION: The Pix-2-Pix GAN model for generating AC-PET demonstrates SUV metrics that highly correlate with original images. AI-generated PET images show clinical potential for reducing the need for CT scans for attenuation correction while preserving quantitative markers and image quality.

Evaluation of a Cascaded Deep Learning-based Algorithm for Prostate Lesion Detection at Biparametric MRI.

Background Multiparametric MRI (mpMRI) improves prostate cancer (PCa) detection compared with systematic biopsy, but its interpretation is prone to interreader variation, which results in performance inconsistency. Artificial intelligence (AI) models can assist in mpMRI interpretation, but large training data sets and extensive model testing are required. Purpose To evaluate a biparametric MRI AI algorithm for intraprostatic lesion detection and segmentation and to compare its performance with radiologist readings and biopsy results. Materials and Methods This secondary analysis of a prospective registry included consecutive patients with suspected or known PCa who underwent mpMRI, US-guided systematic biopsy, or combined systematic and MRI/US fusion-guided biopsy between April 2019 and September 2022. All lesions were prospectively evaluated using Prostate Imaging Reporting and Data System version 2.1. The lesion- and participant-level performance of a previously developed cascaded deep learning algorithm was compared with histopathologic outcomes and radiologist readings using sensitivity, positive predictive value (PPV), and Dice similarity coefficient (DSC). Results A total of 658 male participants (median age, 67 years [IQR, 61-71 years]) with 1029 MRI-visible lesions were included. At histopathologic analysis, 45% (294 of 658) of participants had lesions of International Society of Urological Pathology (ISUP) grade group (GG) 2 or higher. The algorithm identified 96% (282 of 294; 95% CI: 94%, 98%) of all participants with clinically significant PCa, whereas the radiologist identified 98% (287 of 294; 95% CI: 96%, 99%; P = .23). The algorithm identified 84% (103 of 122), 96% (152 of 159), 96% (47 of 49), 95% (38 of 40), and 98% (45 of 46) of participants with ISUP GG 1, 2, 3, 4, and 5 lesions, respectively. In the lesion-level analysis using radiologist ground truth, the detection sensitivity was 55% (569 of 1029; 95% CI: 52%, 58%), and the PPV was 57% (535 of 934; 95% CI: 54%, 61%). The mean number of false-positive lesions per participant was 0.61 (range, 0-3). The lesion segmentation DSC was 0.29. Conclusion The AI algorithm detected cancer-suspicious lesions on biparametric MRI scans with a performance comparable to that of an experienced radiologist. Moreover, the algorithm reliably predicted clinically significant lesions at histopathologic examination. ClinicalTrials.gov Identifier: NCT03354416 © RSNA, 2024 Supplemental material is available for this article.

Automated Detection and Grading of Extraprostatic Extension of Prostate Cancer at MRI via Cascaded Deep Learning and Random Forest Classification.

RATIONALE AND OBJECTIVES: Extraprostatic extension (EPE) is well established as a significant predictor of prostate cancer aggression and recurrence. Accurate EPE assessment prior to radical prostatectomy can impact surgical approach. We aimed to utilize a deep learning-based AI workflow for automated EPE grading from prostate T2W MRI, ADC map, and High B DWI. MATERIAL AND METHODS: An expert genitourinary radiologist conducted prospective clinical assessments of MRI scans for 634 patients and assigned risk for EPE using a grading technique. The training set and held-out independent test set consisted of 507 patients and 127 patients, respectively. Existing deep-learning AI models for prostate organ and lesion segmentation were leveraged to extract area and distance features for random forest classification models. Model performance was evaluated using balanced accuracy, ROC AUCs for each EPE grade, as well as sensitivity, specificity, and accuracy compared to EPE on histopathology. RESULTS: A balanced accuracy score of .390 ± 0.078 was achieved using a lesion detection probability threshold of 0.45 and distance features. Using the test set, ROC AUCs for AI-assigned EPE grades 0-3 were 0.70, 0.65, 0.68, and 0.55 respectively. When using EPE≥ 1 as the threshold for positive EPE, the model achieved a sensitivity of 0.67, specificity of 0.73, and accuracy of 0.72 compared to radiologist sensitivity of 0.81, specificity of 0.62, and accuracy of 0.66 using histopathology as the ground truth. CONCLUSION: Our AI workflow for assigning imaging-based EPE grades achieves an accuracy for predicting histologic EPE approaching that of physicians. This automated workflow has the potential to enhance physician decision-making for assessing the risk of EPE in patients undergoing treatment for prostate cancer due to its consistency and automation.

Reducing False-Positives Due to Urinary Stagnation in the Prostatic Urethra on 18 F-DCFPyL PSMA PET/CT With MRI.

PURPOSE: Prostate-specific membrane antigen (PSMA)-targeting PET radiotracers reveal physiologic uptake in the urinary system, potentially misrepresenting activity in the prostatic urethra as an intraprostatic lesion. This study examined the correlation between midline 18 F-DCFPyL activity in the prostate and hyperintensity on T2-weighted (T2W) MRI as an indication of retained urine in the prostatic urethra. PATIENTS AND METHODS: Eighty-five patients who underwent both 18 F-DCFPyL PSMA PET/CT and prostate MRI between July 2017 and September 2023 were retrospectively analyzed for midline radiotracer activity and retained urine on postvoid T2W MRIs. Fisher's exact tests and unpaired t tests were used to compare residual urine presence and prostatic urethra measurements between patients with and without midline radiotracer activity. The influence of anatomical factors including prostate volume and urethral curvature on urinary stagnation was also explored. RESULTS: Midline activity on PSMA PET imaging was seen in 14 patients included in the case group, whereas the remaining 71 with no midline activity constituted the control group. A total of 71.4% (10/14) and 29.6% (21/71) of patients in the case and control groups had urethral hyperintensity on T2W MRI, respectively ( P < 0.01). Patients in the case group had significantly larger mean urethral dimensions, larger prostate volumes, and higher incidence of severe urethral curvature compared with the controls. CONCLUSIONS: Stagnated urine within the prostatic urethra is a potential confounding factor on PSMA PET scans. Integrating PET imaging with T2W MRI can mitigate false-positive calls, especially as PSMA PET/CT continues to gain traction in diagnosing localized prostate cancer.

Evaluating Diagnostic Accuracy and Inter-reader Agreement of the Prostate Imaging After Focal Ablation Scoring System.

Background and objective: Focal therapy (FT) is increasingly recognized as a promising approach for managing localized prostate cancer (PCa), notably reducing treatment-related morbidities. However, post-treatment anatomical changes present significant challenges for surveillance using current imaging techniques. This study aimed to evaluate the inter-reader agreement and efficacy of the Prostate Imaging after Focal Ablation (PI-FAB) scoring system in detecting clinically significant prostate cancer (csPCa) on post-FT multiparametric magnetic resonance imaging (mpMRI). Methods: A retrospective cohort study was conducted involving patients who underwent primary FT for localized csPCa between 2013 and 2023, followed by post-FT mpMRI and a prostate biopsy. Two expert genitourinary radiologists retrospectively evaluated post-FT mpMRI using PI-FAB. The key measures included inter-reader agreement of PI-FAB scores, assessed by quadratic weighted Cohen's kappa (κ), and the system's efficacy in predicting in-field recurrence of csPCa, with a PI-FAB score cutoff of 3. Additional diagnostic metrics including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy were also evaluated. Key findings and limitations: Scans from 38 patients were analyzed, revealing a moderate level of agreement in PI-FAB scoring (κ = 0.56). Both radiologists achieved sensitivity of 93% in detecting csPCa, although specificity, PPVs, NPVs, and accuracy varied. Conclusions and clinical implications: The PI-FAB scoring system exhibited high sensitivity with moderate inter-reader agreement in detecting in-field recurrence of csPCa. Despite promising results, its low specificity and PPV necessitate further refinement. These findings underscore the need for larger studies to validate the clinical utility of PI-FAB, potentially aiding in standardizing post-treatment surveillance. Patient summary: Focal therapy has emerged as a promising approach for managing localized prostate cancer, but limitations in current imaging techniques present significant challenges for post-treatment surveillance. The Prostate Imaging after Focal Ablation (PI-FAB) scoring system showed high sensitivity for detecting in-field recurrence of clinically significant prostate cancer. However, its low specificity and positive predictive value necessitate further refinement. Larger, more comprehensive studies are needed to fully validate its clinical utility.

PRECISE Version 2: Updated Recommendations for Reporting Prostate Magnetic Resonance Imaging in Patients on Active Surveillance for Prostate Cancer.

BACKGROUND AND OBJECTIVE: The Prostate Cancer Radiological Estimation of Change in Sequential Evaluation (PRECISE) recommendations standardise the reporting of prostate magnetic resonance imaging (MRI) in patients on active surveillance (AS) for prostate cancer. An international consensus group recently updated these recommendations and identified the areas of uncertainty. METHODS: A panel of 38 experts used the formal RAND/UCLA Appropriateness Method consensus methodology. Panellists scored 193 statements using a 1-9 agreement scale, where 9 means full agreement. A summary of agreement, uncertainty, or disagreement (derived from the group median score) and consensus (determined using the Interpercentile Range Adjusted for Symmetry method) was calculated for each statement and presented for discussion before individual rescoring. KEY FINDINGS AND LIMITATIONS: Participants agreed that MRI scans must meet a minimum image quality standard (median 9) or be given a score of 'X' for insufficient quality. The current scan should be compared with both baseline and previous scans (median 9), with the PRECISE score being the maximum from any lesion (median 8). PRECISE 3 (stable MRI) was subdivided into 3-V (visible) and 3-NonV (nonvisible) disease (median 9). Prostate Imaging Reporting and Data System/Likert ≥3 lesions should be measured on T2-weighted imaging, using other sequences to aid in the identification (median 8), and whenever possible, reported pictorially (diagrams, screenshots, or contours; median 9). There was no consensus on how to measure tumour size. More research is needed to determine a significant size increase (median 9). PRECISE 5 was clarified as progression to stage ≥T3a (median 9). CONCLUSIONS AND CLINICAL IMPLICATIONS: The updated PRECISE recommendations reflect expert consensus opinion on minimal standards and reporting criteria for prostate MRI in AS. PATIENT SUMMARY: The Prostate Cancer Radiological Estimation of Change in Sequential Evaluation (PRECISE) recommendations are used in clinical practice and research to guide the interpretation and reporting of magnetic resonance imaging for patients on active surveillance for prostate cancer. An international panel has updated these recommendations, clarified the areas of uncertainty, and highlighted the areas for further research.

Automated prostate gland segmentation in challenging clinical cases: comparison of three artificial intelligence methods.

OBJECTIVE: Automated methods for prostate segmentation on MRI are typically developed under ideal scanning and anatomical conditions. This study evaluates three different prostate segmentation AI algorithms in a challenging population of patients with prior treatments, variable anatomic characteristics, complex clinical history, or atypical MRI acquisition parameters. MATERIALS AND METHODS: A single institution retrospective database was queried for the following conditions at prostate MRI: prior prostate-specific oncologic treatment, transurethral resection of the prostate (TURP), abdominal perineal resection (APR), hip prosthesis (HP), diversity of prostate volumes (large ≥ 150 cc, small ≤ 25 cc), whole gland tumor burden, magnet strength, noted poor quality, and various scanners (outside/vendors). Final inclusion criteria required availability of axial T2-weighted (T2W) sequence and corresponding prostate organ segmentation from an expert radiologist. Three previously developed algorithms were evaluated: (1) deep learning (DL)-based model, (2) commercially available shape-based model, and (3) federated DL-based model. Dice Similarity Coefficient (DSC) was calculated compared to expert. DSC by model and scan factors were evaluated with Wilcox signed-rank test and linear mixed effects (LMER) model. RESULTS: 683 scans (651 patients) met inclusion criteria (mean prostate volume 60.1 cc [9.05-329 cc]). Overall DSC scores for models 1, 2, and 3 were 0.916 (0.707-0.971), 0.873 (0-0.997), and 0.894 (0.025-0.961), respectively, with DL-based models demonstrating significantly higher performance (p < 0.01). In sub-group analysis by factors, Model 1 outperformed Model 2 (all p < 0.05) and Model 3 (all p < 0.001). Performance of all models was negatively impacted by prostate volume and poor signal quality (p < 0.01). Shape-based factors influenced DL models (p < 0.001) while signal factors influenced all (p < 0.001). CONCLUSION: Factors affecting anatomical and signal conditions of the prostate gland can adversely impact both DL and non-deep learning-based segmentation models.

Localized high-risk prostate cancer harbors an androgen receptor low subpopulation susceptible to HER2 inhibition.

Patients diagnosed with localized high-risk prostate cancer have higher rates of recurrence, and the introduction of neoadjuvant intensive hormonal therapies seeks to treat occult micrometastatic disease by their addition to definitive treatment. Sufficient profiling of baseline disease has remained a challenge in enabling the in-depth assessment of phenotypes associated with exceptional vs. poor pathologic responses after treatment. In this study, we report comprehensive and integrative gene expression profiling of 37 locally advanced prostate tumors prior to six months of androgen deprivation therapy (ADT) plus the androgen receptor (AR) inhibitor enzalutamide prior to radical prostatectomy. A robust transcriptional program associated with HER2 activity was positively associated with poor outcome and opposed AR activity, even after adjusting for common genomic alterations in prostate cancer including PTEN loss and expression of the TMPRSS2:ERG fusion. Patients experiencing exceptional pathologic responses demonstrated lower levels of HER2 and phospho-HER2 by immunohistochemistry of biopsy tissues. The inverse correlation of AR and HER2 activity was found to be a universal feature of all aggressive prostate tumors, validated by transcriptional profiling an external cohort of 121 patients and immunostaining of tumors from 84 additional patients. Importantly, the AR activity-low, HER2 activity-high cells that resist ADT are a pre-existing subset of cells that can be targeted by HER2 inhibition alone or in combination with enzalutamide. In summary, we show that prostate tumors adopt an AR activity-low prior to antiandrogen exposure that can be exploited by treatment with HER2 inhibitors.

Deep Learning-Based Detection and Classification of Bone Lesions on Staging Computed Tomography in Prostate Cancer: A Development Study.

RATIONALE AND OBJECTIVES: Efficiently detecting and characterizing metastatic bone lesions on staging CT is crucial for prostate cancer (PCa) care. However, it demands significant expert time and additional imaging such as PET/CT. We aimed to develop an ensemble of two automated deep learning AI models for 1) bone lesion detection and segmentation and 2) benign vs. metastatic lesion classification on staging CTs and to compare its performance with radiologists. MATERIALS AND METHODS: This retrospective study developed two AI models using 297 staging CT scans (81 metastatic) with 4601 benign and 1911 metastatic lesions in PCa patients. Metastases were validated by follow-up scans, bone biopsy, or PET/CT. Segmentation AI (3DAISeg) was developed using the lesion contours delineated by a radiologist. 3DAISeg performance was evaluated with the Dice similarity coefficient, and classification AI (3DAIClass) performance on AI and radiologist contours was assessed with F1-score and accuracy. Training/validation/testing data partitions of 70:15:15 were used. A multi-reader study was performed with two junior and two senior radiologists within a subset of the testing dataset (n = 36). RESULTS: In 45 unseen staging CT scans (12 metastatic PCa) with 669 benign and 364 metastatic lesions, 3DAISeg detected 73.1% of metastatic (266/364) and 72.4% of benign lesions (484/669). Each scan averaged 12 extra segmentations (range: 1-31). All metastatic scans had at least one detected metastatic lesion, achieving a 100% patient-level detection. The mean Dice score for 3DAISeg was 0.53 (median: 0.59, range: 0-0.87). The F1 for 3DAIClass was 94.8% (radiologist contours) and 92.4% (3DAISeg contours), with a median false positive of 0 (range: 0-3). Using radiologist contours, 3DAIClass had PPV and NPV rates comparable to junior and senior radiologists: PPV (semi-automated approach AI 40.0% vs. Juniors 32.0% vs. Seniors 50.0%) and NPV (AI 96.2% vs. Juniors 95.7% vs. Seniors 91.9%). When using 3DAISeg, 3DAIClass mimicked junior radiologists in PPV (pure-AI 20.0% vs. Juniors 32.0% vs. Seniors 50.0%) but surpassed seniors in NPV (pure-AI 93.8% vs. Juniors 95.7% vs. Seniors 91.9%). CONCLUSION: Our lesion detection and classification AI model performs on par with junior and senior radiologists in discerning benign and metastatic lesions on staging CTs obtained for PCa.

AI-powered radiomics: revolutionizing detection of urologic malignancies.

PURPOSE OF REVIEW: This review aims to highlight the integration of artificial intelligence-powered radiomics in urologic oncology, focusing on the diagnostic and prognostic advancements in the realm of managing prostate, kidney, and bladder cancers. RECENT FINDINGS: As artificial intelligence continues to shape the medical imaging landscape, its integration into the field of urologic oncology has led to impressive results. For prostate cancer diagnostics, machine learning has shown promise in refining clinically-significant lesion detection, with some success in deciphering ambiguous lesions on multiparametric MRI. For kidney cancer, radiomics has emerged as a valuable tool for better distinguishing between benign and malignant renal masses and predicting tumor behavior from CT or MRI scans. Meanwhile, in the arena of bladder cancer, there is a burgeoning emphasis on prediction of muscle invasive cancer and forecasting disease trajectory. However, many studies showing promise in these areas face challenges due to limited sample sizes and the need for broader external validation. SUMMARY: Radiomics integrated with artificial intelligence offers a pioneering approach to urologic oncology, ushering in an era of enhanced diagnostic precision and reduced invasiveness, guiding patient-tailored treatment plans. Researchers must embrace broader, multicentered endeavors to harness the full potential of this field.

Quality of T2-weighted MRI re-acquisition versus deep learning GAN image reconstruction: A multi-reader study.

PURPOSE: To evaluate CycleGAN's ability to enhance T2-weighted image (T2WI) quality. METHOD: A CycleGAN algorithm was used to enhance T2WI quality. 96 patients (192 scans) were identified from patients who underwent multiple axial T2WI due to poor quality on the first attempt (RAD1) and improved quality on re-acquisition (RAD2). CycleGAN algorithm gave DL classifier scores (0-1) for quality quantification and produced enhanced versions of QI1 and QI2 from RAD1 and RAD2, respectively. A subset (n = 20 patients) was selected for a blinded, multi-reader study, where four radiologists rated T2WI on a scale of 1-4 for quality. The multi-reader study presented readers with 60 image pairs (RAD1 vs RAD2, RAD1 vs QI1, and RAD2 vs QI2), allowing for selecting sequence preferences and quantifying the quality changes. RESULTS: The DL classifier correctly discerned 71.9 % of quality classes, with 90.6 % (96/106) as poor quality and 48.8 % (42/86) as diagnostic in original sequences (RAD1, RAD2). CycleGAN images (QI1, QI2) demonstrated quantitative improvements, with consistently higher DL classifier scores than original scans (p < 0.001). In the multi-reader analysis, CycleGAN demonstrated no qualitative improvements, with diminished overall quality and motion in QI2 in most patients compared to RAD2, with noise levels remaining similar (8/20). No readers preferred QI2 to RAD2 for diagnosis. CONCLUSION: Despite quantitative enhancements with CycleGAN, there was no qualitative boost in T2WI diagnostic quality, noise, or motion. Expert radiologists didn't favor CycleGAN images over standard scans, highlighting the divide between quantitative and qualitative metrics.