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 case series of osseous metastases in patients with glioblastoma.

BACKGROUND: Extracranial metastases occur in <2% of cases of glioblastoma (GBM). When metastases do occur, bone is the most common destination. Herein, we review clinical characteristics of GBM patients with osseous metastases and evaluate both potential risk factors and prognostic significance. METHODS: Using an institutional database, we identified and retrospectively analyzed 6 patients with both GBM and osseous metastases. We collected data on patient demographics, tumor genetics, clinical courses, and outcomes. Given the rarity of metastatic GBM, we conducted historical comparisons using previously published literature. RESULTS: Five patients with osseous metastases (83%) were male, with a median age of 46 years at GBM diagnosis (range: 20-84). All patients had IDH-wildtype, MGMT promoter unmethylated GBM and 5 (83%) had alterations in TP53. All patients underwent surgical resection for GBM followed by radiation with concurrent and adjuvant temozolomide. Four patients (67%) received bevacizumab prior to bone metastasis diagnosis. Bone metastases were discovered at a median of 12.2 months (range: 5.3-35.2) after GBM diagnosis and 4.8 months after starting bevacizumab (range: 3.5-13.2). Three patients (50%) received immunotherapy. After osseous metastasis diagnosis, the median survival was 25 days (range: 13-225). CONCLUSION: In our cohort, most patients were male and young at the time of GBM diagnosis. All patients had IDH-wildtype, MGMT promoter unmethylated GBM, and most had alterations in TP53, which may be important for osseous metastasis. Most patients received bevacizumab, which has been associated with earlier metastasis. Osseous metastases of GBM occur and portend a dismal prognosis in an already aggressive malignancy.

Expression of tumor antigens within an oncolytic virus enhances the anti-tumor T cell response.

Although patients benefit from immune checkpoint inhibition (ICI) therapy in a broad variety of tumors, resistance may arise from immune suppressive tumor microenvironments (TME), which is particularly true of hepatocellular carcinoma (HCC). Since oncolytic viruses (OV) can generate a highly immune-infiltrated, inflammatory TME, OVs could potentially restore ICI responsiveness via recruitment, priming, and activation of anti-tumor T cells. Here we find that on the contrary, an oncolytic vesicular stomatitis virus, expressing interferon-ß (VSV-IFNß), antagonizes the effect of anti-PD-L1 therapy in a partially anti-PD-L1-responsive model of HCC. Cytometry by Time of Flight shows that VSV-IFNß expands dominant anti-viral effector CD8 T cells with concomitant relative disappearance of anti-tumor T cell populations, which are the target of anti-PD-L1. However, by expressing a range of HCC tumor antigens within VSV, combination OV and anti-PD-L1 therapeutic benefit could be restored. Our data provide a cautionary message for the use of highly immunogenic viruses as tumor-specific immune-therapeutics by showing that dominant anti-viral T cell responses can inhibit sub-dominant anti-tumor T cell responses. However, through encoding tumor antigens within the virus, oncolytic virotherapy can generate anti-tumor T cell populations upon which immune checkpoint blockade can effectively work.

3-Position Biaryl Endochin-like Quinolones with Enhanced Antimalarial Performance.

ELQ-300 is a potent antimalarial drug with activity against blood, liver, and vector stages of the disease. A prodrug, ELQ-331, exhibits reduced crystallinity and improved in vivo efficacy in preclinical testing, and currently, it is in the developmental pipeline for once-a-week dosing for oral prophylaxis against malaria. Because of the high cost of developing a new drug for human use and the high risk of drug failure, it is prudent to have a back-up plan in place. Here we describe ELQ-596, a member of a new subseries of 3-biaryl-ELQs, with enhanced potency in vitro against multidrug-resistant Plasmodium falciparum parasites. ELQ-598, a prodrug of ELQ-596 with diminished crystallinity, is more effective vs murine malaria than its progenitor ELQ-331 by 4- to 10-fold, suggesting that correspondingly lower doses could be used to protect and cure humans of malaria. With a longer bloodstream half-life in mice compared to its progenitor, ELQ-596 highlights a novel series of next-generation ELQs with the potential for once-monthly dosing for protection against malaria infection. Advances in the preparation of 3-biaryl-ELQs are presented along with preliminary results from experiments to explore key structure-activity relationships for drug potency, selectivity, pharmacokinetics, and safety.

Comparison of Artificial Intelligence to Resident Performance on Upper-Extremity Orthopaedic In-Training Examination Questions.

Purpose: Currently, there is a paucity of prior investigations and studies examining applications for artificial intelligence (AI) in upper-extremity (UE) surgical education. The purpose of this investigation was to assess the performance of a novel AI tool (ChatGPT) on UE questions on the Orthopaedic In-Training Examination (OITE). We aimed to compare the performance of ChatGPT to the examination performance of hand surgery residents. Methods: We selected questions from the 2020-2022 OITEs that focused on both the hand and UE as well as the shoulder and elbow content domains. These questions were divided into two categories: those with text-only prompts (text-only questions) and those that included supplementary images or videos (media questions). Two authors (B.K.F. and G.S.M.) converted the accompanying media into text-based descriptions. Included questions were inputted into ChatGPT (version 3.5) to generate responses. Each OITE question was entered into ChatGPT three times: (1) open-ended response, which requested a free-text response; (2) multiple-choice responses without asking for justification; and (3) multiple-choice response with justification. We referred to the OITE scoring guide for each year in order to compare the percentage of correct AI responses to correct resident responses. Results: A total of 102 UE OITE questions were included; 59 were text-only questions, and 43 were media-based. ChatGPT correctly answered 46 (45%) of 102 questions using the Multiple Choice No Justification prompt requirement (42% for text-based and 44% for media questions). Compared to ChatGPT, postgraduate year 1 orthopaedic residents achieved an average score of 51% correct. Postgraduate year 5 residents answered 76% of the same questions correctly. Conclusions: ChatGPT answered fewer UE OITE questions correctly compared to hand surgery residents of all training levels. Clinical relevance: Further development of novel AI tools may be necessary if this technology is going to have a role in UE education.

Variations in germ cell tumor histology by age and implications for cancer-specific survival among pediatric and adult males: A population-based study.

PURPOSE: Few studies have quantified differences in histology and implications for survival between male children and adults with germ cell tumors (GCT). We evaluated these differences and associations with cancer-specific survival (CSS) using Surveillance, Epidemiology, and End Results (SEER) cancer registries. METHODS: SEER (1988-2016) was used to identify male patients 0 to 40 years of age diagnosed with seminoma and nonseminomatous GCT (NSGCT). Demographic and tumor characteristics were tabulated with histology distributions compared by age group (0-4, 12-18, 19-40 years old). CSS was evaluated in multivariable Cox proportional hazards regression models. RESULTS: Among 27,204 patients identified, 1,538 (5.7%) were pediatric (0-18 years). Seminoma (54.3%) predominated in adult patients (ages 19-40). Among 0 to 4 years-old, yolk sac tumor (71.2%) and teratoma (21.5%) were most common. Mixed GCT (52.7%) was most prevalent among 12 to 18 years-old with seminoma, embryonal, and teratoma occurring in 12 to 15% each. Relative to pediatric patients, adult patients had similar CSS for seminoma but worse CSS for NSGCT on Kaplan-Meier curves with 9 years mean follow-up. Choriocarcinoma and yolk sac tumors carried the worst prognosis relative to seminoma for both children (HR 5.7 and HR 11.1, respectively, both P < 0.01) and adults (HR 4.6 and HR 4.6, respectively, both P < 0.01) adjusted for stage. CONCLUSION: Histology of GCTs vary by age with yolk sac tumors and teratoma predominating for male patients 0 to 4 years, mixed GCT for 12 to 18 years, and seminoma for 19 to 40 years. Pediatric patients with NSGCT had higher CSS than their adult counterparts. Mixed GCT represented an increasing proportion of GCT over the study period. Age, stage, and histology impact CSS in both pediatric and adult populations.

Spontaneous and double-strand break repair-associated quasipalindrome and frameshift mutagenesis in budding yeast: role of mismatch repair.

Although gene conversion (GC) in Saccharomyces cerevisiae is the most error-free way to repair double-strand breaks (DSBs), the mutation rate during homologous recombination is 1,000 times greater than during replication. Many mutations involve dissociating a partially copied strand from its repair template and re-aligning with the same or another template, leading to -1 frameshifts in homonucleotide runs, quasipalindrome (QP)-associated mutations and microhomology-mediated interchromosomal template switches. We studied GC induced by HO endonuclease cleavage at MATα, repaired by an HMR::KI-URA3 donor. We inserted into HMR::KI-URA3 an 18-bp inverted repeat where one arm had a 4-bp insertion. Most GCs yield MAT::KI-ura3::QP + 4 (Ura-) outcomes, but template-switching produces Ura+ colonies, losing the 4-bp insertion. If the QP arm without the insertion is first encountered by repair DNA polymerase and is then (mis)used as a template, the palindrome is perfected. When the QP + 4 arm is encountered first, Ura+ derivatives only occur after second-end capture and second-strand synthesis. QP + 4 mutations are suppressed by mismatch repair (MMR) proteins Msh2, Msh3, and Mlh1, but not Msh6. Deleting Rdh54 significantly reduces QP mutations only when events creating Ura+ occur in the context of a D-loop but not during second-strand synthesis. A similar bias is found with a proofreading-defective DNA polymerase mutation (poI3-01). DSB-induced mutations differed in several genetic requirements from spontaneous events. We also created a + 1 frameshift in the donor, expanding a run of 4 Cs to 5 Cs. Again, Ura+ recombinants markedly increased by disabling MMR, suggesting that MMR acts during GC but favors the unbroken, template strand.

Immune Checkpoint Inhibitors and Glioblastoma: A Review on Current State and Future Directions.

Glioblastoma (GBM) is the most prevalent malignant tumor of the central nervous system. The prognosis of GBM is grim, with a median overall survival of 14.6 months and only 6.9% of patients surviving 5 years after the initial diagnosis. Despite poor outcomes, standard therapy of surgical resection, radiotherapy, chemotherapy, and tumor-treating fields has remained largely unchanged. The introduction of immune checkpoint inhibitors (ICI) has been a paradigm shift in oncology, with efficacy across a broad spectrum of cancer types. Nonetheless, investigations of ICIs in both newly diagnosed and recurrent GBM have thus far been disappointing. This lack of clinical benefit has been largely attributed to the highly immunosuppressive nature of GBM. However, immunotherapy still holds promise for the treatment of GBM, with combinatorial strategies offering hope for potentially overcoming these current limitations. In this review, we discuss the outcomes of clinical trials employing ICIs in patients with GBM. Afterward, we review ICI combination strategies and how these combinations may overcome the immunosuppressive microenvironment of GBM in the context of preclinical/clinical evidence and ongoing clinical trials.

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.

Ghost roads and the destruction of Asia-Pacific tropical forests.

Roads are expanding at the fastest pace in human history. This is the case especially in biodiversity-rich tropical nations, where roads can result in forest loss and fragmentation, wildfires, illicit land invasions and negative societal effects1-5. Many roads are being constructed illegally or informally and do not appear on any existing road map6-10; the toll of such 'ghost roads' on ecosystems is poorly understood. Here we use around 7,000 h of effort by trained volunteers to map ghost roads across the tropical Asia-Pacific region, sampling 1.42 million plots, each 1 km2 in area. Our intensive sampling revealed a total of 1.37 million km of roads in our plots-from 3.0 to 6.6 times more roads than were found in leading datasets of roads globally. Across our study area, road building almost always preceded local forest loss, and road density was by far the strongest correlate11 of deforestation out of 38 potential biophysical and socioeconomic covariates. The relationship between road density and forest loss was nonlinear, with deforestation peaking soon after roads penetrate a landscape and then declining as roads multiply and remaining accessible forests largely disappear. Notably, after controlling for lower road density inside protected areas, we found that protected areas had only modest additional effects on preventing forest loss, implying that their most vital conservation function is limiting roads and road-related environmental disruption. Collectively, our findings suggest that burgeoning, poorly studied ghost roads are among the gravest of all direct threats to tropical forests.

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.

Cerebrospinal Fluid Liquid Biopsies in the Evaluation of Adult Gliomas.

PURPOSE OF REVIEW: This review aims to discuss recent research regarding the biomolecules explored in liquid biopsies and their potential clinical uses for adult-type diffuse gliomas. RECENT FINDINGS: Evaluation of tumor biomolecules via cerebrospinal fluid (CSF) is an emerging technology in neuro-oncology. Studies to date have already identified various circulating tumor DNA, extracellular vesicle, micro-messenger RNA and protein biomarkers of interest. These biomarkers show potential to assist in multiple avenues of central nervous system (CNS) tumor evaluation, including tumor differentiation and diagnosis, treatment selection, response assessment, detection of tumor progression, and prognosis. In addition, CSF liquid biopsies have the potential to better characterize tumor heterogeneity compared to conventional tissue collection and CNS imaging. Current imaging modalities are not sufficient to establish a definitive glioma diagnosis and repeated tissue sampling via conventional biopsy is risky, therefore, there is a great need to improve non-invasive and minimally invasive sampling methods. CSF liquid biopsies represent a promising, minimally invasive adjunct to current approaches which can provide diagnostic and prognostic information as well as aid in response assessment.

Global dominance of lianas over trees is driven by forest disturbance, climate and topography.

Growing evidence suggests that liana competition with trees is threatening the global carbon sink by slowing the recovery of forests following disturbance. A recent theory based on local and regional evidence further proposes that the competitive success of lianas over trees is driven by interactions between forest disturbance and climate. We present the first global assessment of liana-tree relative performance in response to forest disturbance and climate drivers. Using an unprecedented dataset, we analysed 651 vegetation samples representing 26,538 lianas and 82,802 trees from 556 unique locations worldwide, derived from 83 publications. Results show that lianas perform better relative to trees (increasing liana-to-tree ratio) when forests are disturbed, under warmer temperatures and lower precipitation and towards the tropical lowlands. We also found that lianas can be a critical factor hindering forest recovery in disturbed forests experiencing liana-favourable climates, as chronosequence data show that high competitive success of lianas over trees can persist for decades following disturbances, especially when the annual mean temperature exceeds 27.8°C, precipitation is less than 1614 mm and climatic water deficit is more than 829 mm. These findings reveal that degraded tropical forests with environmental conditions favouring lianas are disproportionately more vulnerable to liana dominance and thus can potentially stall succession, with important implications for the global carbon sink, and hence should be the highest priority to consider for restoration management.

Des preuves de plus en plus nombreuses suggèrent que la competition entre lianes et les arbres menace le puits de carbone mondial en ralentissant la récupération des forêts après une perturbation. Une théorie récente, fondée sur des observations locales et régionales, propose en outre que le succès compétitif des lianes sur les arbres est dû aux interactions entre la perturbation forestière et le climat. Nous présentons la première évaluation mondiale de la performance relative des lianes par rapport aux arbres en réponse aux perturbations forestières et aux facteurs climatiques. En utilisant un ensemble de données sans précédent, nous avons analysé 651 échantillons de végétation représentant 26,538 lianes et 82,802 arbres, issus de 556 emplacements uniques dans le monde entier, tirés de 83 publications. Les résultats montrent que les lianes ont de meilleure performances par rapport aux arbres (augmentation du ratio liane-arbre) lorsque les forêts sont perturbées, sous des zones chaudes aves précipitations faibles, et vers les basses altitudes tropicales. Nous avons également constaté que les lianes peuvent être un facteur critique entravant la récupération des forêts dans les forêts perturbées connaissant des climats favorables aux lianes, car les données de chronoséquence montrent que le succès compétitif élevé des lianes sur les arbres peut persister pendant des décennies après les perturbations, surtout lorsque la température annuelle moyenne dépasse 27.8°C, que les précipitations sont inférieures à 1614 mm et que le déficit hydrique climatique est supérieur à 829 mm. Ces découvertes révèlent que les forêts tropicales dégradées avec des conditions environnementales favorables aux lianes sont disproportionnellement plus vulnérables à la dominance des lianes, et peuvent ainsi potentiellement entraver la succession, avec d'importantes implications pour le puits de carbone mondial et devraient donc être la plus haute priorité à considérer pour la gestion de la restauration.

Long-term outcomes of pelvic-fascia sparing robotic-assisted radical prostatectomy versus standard technique: Superior urinary function and quality of life without compromising oncologic efficacy in a single-surgeon series.

BACKGROUND: Prostatic fascia-sparing robotic-assisted radical prostatectomy (PFS-RARP) has improved short-term postoperative continence compared to standard prostatectomy (S-RARP) but long-term differences remain unclear. MATERIALS AND METHODS: One hundred two S-RARP followed by 239 PFS-RARPs were performed by a single surgeon. Univariate analyses were performed with t-test, χ2, Wilcoxon rank sum, Fisher exact, and analysis of variance (ANOVA). Regression models analyzed associates of EPIC-CP scores and oncologic outcomes. Cox proportional hazards modeling assessed postoperative continence. Primary outcomes included patient-reported urinary incontinence (UI) via EPIC-CP and continence rates. Secondary outcomes included EPIC-CP scores, positive surgical margins (PSM), and biochemical recurrence (BCR). Perioperative outcomes and time to continence were measured. RESULTS: Median follow-up for PFS-RARP vs. S-RARP was 26 vs. 65 months. PFS-RARP demonstrated improved EPIC-CP UI and total scores at 24 months. On multivariate analysis, PFS-RARP was associated with improved EPIC-CP UI and total scores through 18 months, but not with PSM or BCR. PFS-RARP had a 39% and 66% reduced risk of incontinence using 0 and 0 to 1 pad-use definitions (HR 0.61, 95% CI 0.39 - 0.95; HR:0.34, 95% CI 0.16 - 0.76). Continence returned faster with PFS-RARP (0 PPD: 91.0 days vs. 261 days, P < 0.001; 0-1 PPD: 32.7 days vs. 171 days, P < 0.001). There were no differences in PSM (35% vs. 25%, P = 0.064). There were more anterior PSM in PFS-RARP vs. S-RARP (47% vs. 26% P = 0.035), but no differences in BCR (16% vs. 22% P = 0.241). CONCLUSIONS: PFS-RARP improves continence and patient-reported QOL up to 24 months postoperatively without compromising oncologic outcomes.

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.

Exploring Trends in Immediate Postresidency Career Paths: A Multi-year Analysis of Plastic Surgery Resident Graduates Across the United States.

Background: Plastic surgery is one of the most diverse specialties in medicine. Because of the competitiveness of plastic surgery residency, applicants are entering the field with increased experience and more developed interests in specific specialties. Programs and prospective applicants may find it beneficial to know trends in the career paths of recent graduates. Objectives: To identify trends in postresidency career paths for plastic surgery graduates. Methods: Data from all integrated plastic surgery residency programs were analyzed from 2013 to 2022. Eighty-eight residency programs were analyzed for review. Residency websites were the primary source of data. Postresidency career paths were categorized into subspecialty fellowships, academic practice, or private practice. Secondary data included program rank, size of the program, associated fellowship program, associated independent program, and program location. Results: Seventy-three programs met the inclusion criteria. Private practice was the most common immediate postgraduation path. Microvascular and aesthetic fellowships demonstrated maximum growth in the last 10 years, followed by hand fellowships. Programs ranked in the top 25 by Doximity reputation were significantly associated with graduates going into craniofacial (P = .05) and microvascular fellowship (P = .021), and immediate academic practice (P = .011). Lower-ranked programs were correlated with higher levels of graduates entering directly into private/community hospital practice (ρ = 0.327). Conclusions: Life after residency is a necessary consideration for training physicians. Understanding trends in postresidency career paths could help programs and prospective applicants make more informed decisions on what programs may offer the best opportunities to pursue their desired career path.

PI-RADS Version 2.0 Versus Version 2.1: Comparison of Prostate Cancer Gleason Grade Upgrade and Downgrade Rates From MRI-Targeted Biopsy to Radical Prostatectomy.

BACKGROUND. Precise risk stratification through MRI/ultrasound (US) fusion-guided targeted biopsy (TBx) can guide optimal prostate cancer (PCa) management. OBJECTIVE. The purpose of this study was to compare PI-RADS version 2.0 (v2.0) and PI-RADS version 2.1 (v2.1) in terms of the rates of International Society of Urological Pathology (ISUP) grade group (GG) upgrade and downgrade from TBx to radical prostatectomy (RP). METHODS. This study entailed a retrospective post hoc analysis of patients who underwent 3-T prostate MRI at a single institution from May 2015 to March 2023 as part of three prospective clinical trials. Trial participants who underwent MRI followed by MRI/US fusion-guided TBx and RP within a 1-year interval were identified. A single genitourinary radiologist performed clinical interpretations of the MRI examinations using PI-RADS v2.0 from May 2015 to March 2019 and PI-RADS v2.1 from April 2019 to March 2023. Upgrade and downgrade rates from TBx to RP were compared using chi-square tests. Clinically significant cancer was defined as ISUP GG2 or greater. RESULTS. The final analysis included 308 patients (median age, 65 years; median PSA density, 0.16 ng/mL2). The v2.0 group (n = 177) and v2.1 group (n = 131) showed no significant difference in terms of upgrade rate (29% vs 22%, respectively; p = .15), downgrade rate (19% vs 21%, p = .76), clinically significant upgrade rate (14% vs 10%, p = .27), or clinically significant downgrade rate (1% vs 1%, p > .99). The upgrade rate and downgrade rate were also not significantly different between the v2.0 and v2.1 groups when stratifying by index lesion PI-RADS category or index lesion zone, as well as when assessed only in patients without a prior PCa diagnosis (all p > .01). Among patients with GG2 or GG3 at RP (n = 121 for v2.0; n = 103 for v2.1), the concordance rate between TBx and RP was not significantly different between the v2.0 and v2.1 groups (53% vs 57%, p = .51). CONCLUSION. Upgrade and downgrade rates from TBx to RP were not significantly different between patients whose MRI examinations were clinically interpreted using v2.0 or v2.1. CLINICAL IMPACT. Implementation of the most recent PI-RADS update did not improve the incongruence in PCa grade assessment between TBx and surgery.

Deep Learning-Based Interpretable AI for Prostate T2W MRI Quality Evaluation.

RATIONALE AND OBJECTIVES: Prostate MRI quality is essential in guiding prostate biopsies. However, assessment of MRI quality is subjective with variation. Quality degradation sources exert varying impacts based on the sequence under consideration, such as T2W versus DWI. As a result, employing sequence-specific techniques for quality assessment could yield more advantageous outcomes. This study aims to develop an AI tool that offers a more consistent evaluation of T2W prostate MRI quality, efficiently identifying suboptimal scans while minimizing user bias. MATERIALS AND METHODS: This retrospective study included 1046 patients from three cohorts (ProstateX [n = 347], All-comer in-house [n = 602], enriched bad-quality MRI in-house [n = 97]) scanned between January 2011 and May 2022. An expert reader assigned T2W MRIs a quality score. A train-validation-test split of 70:15:15 was applied, ensuring equal distribution of MRI scanners and protocols across all partitions. T2W quality AI classification model was based on 3D DenseNet121 architecture using MONAI framework. In addition to multiclassification, binary classification was utilized (Classes 0/1 vs. 2). A score of 0 was given to scans considered non-diagnostic or unusable, a score of 1 was given to those with acceptable diagnostic quality with some usability but with some quality distortions present, and a score of 2 was given to those considered optimal diagnostic quality and usability. Partial occlusion sensitivity maps were generated for anatomical correlation. Three body radiologists assessed reproducibility within a subgroup of 60 test cases using weighted Cohen Kappa. RESULTS: The best validation multiclass accuracy of 77.1% (121/157) was achieved during training. In the test dataset, multiclassification accuracy was 73.9% (116/157), whereas binary accuracy was 84.7% (133/157). Sub-class sensitivity for binary quality distortion classification for class 0 was 100% (18/18), and sub-class specificity for T2W classification of absence/minimal quality distortions for class 2 was 90.5% (95/105). All three readers showed moderate to substantial agreement with ground truth (R1-R3 κ = 0.588, κ = 0.649, κ = 0.487, respectively), moderate to substantial agreement with each other (R1-R2 κ = 0.599, R1-R3 κ = 0.612, R2-R3 κ = 0.685), fair to moderate agreement with AI (R1-R3 κ = 0.445, κ = 0.410, κ = 0.292, respectively). AI showed substantial agreement with ground truth (κ = 0.704). 3D quality heatmap evaluation revealed that the most critical non-diagnostic quality imaging features from an AI perspective related to obscuration of the rectoprostatic space (94.4%, 17/18). CONCLUSION: The 3D AI model can assess T2W prostate MRI quality with moderate accuracy and translate whole sequence-level classification labels into 3D voxel-level quality heatmaps for interpretation. Image quality has a significant downstream impact on ruling out clinically significant cancers. AI may be able to help with reproducible identification of MRI sequences requiring re-acquisition with explainability.

Outcomes and Safety of Suprapubic vs Urethral Catheterization Following Pelvic Fascia‒Sparing Robotic Prostatectomy.

INTRODUCTION: Urethral catheter (UC) discomfort remains a burden following robotic-assisted radical prostatectomy (RARP). Suprapubic catheters (SPCs) may reduce patient discomfort and increase satisfaction. Pelvic fascia‒sparing (PFS) RARP reduces the technical challenges of intraoperative SPC placement. We examined postoperative outcomes of SPC vs UC placement following PFS-RARP. METHODS: We conducted a retrospective review of a prospective institutional review board‒approved database of PFS-RARP patients from June 2020 to December 2022 receiving SPC (n = 108) or UC (n = 104) postoperatively. Demographics and clinical and perioperative outcomes were captured. Postoperative patient-reported quality of life was measured using EPIC-CP (Expanded Prostate Cancer Index Composite for Clinical Practice). Patients with intraoperative complications or intraoperative leaks or undergoing salvage prostatectomy were excluded. Univariate and multivariate regression analyses were performed to compare outcomes. RESULTS: No significant differences in demographics or oncologic outcomes existed. There were no differences in complications, including urethral stricture or anastomotic leak. Men receiving SPC vs UC had earlier return to continence (7 vs 16 days, P < .001) and higher continence rates at catheter removal (67.6% vs 43.3%, P = .0003). On adjusted analyses, SPC was an independent predictor of continence at catheter removal (OR 2.21, P = .023). There were no differences between groups in preoperative or postoperative EPIC-CP scores, including no differences in postoperative quality of life (P = .46). CONCLUSIONS: SPC after PFS-RARP is a safe and feasible alternative to UC. SPC is associated with an earlier return to continence and higher continence rates at catheter removal. Use of SPC may increase overall patient satisfaction following PFS-RARP.