Clinical application of machine learning models in patients with prostate cancer before prostatectomy.

BACKGROUND: To build machine learning predictive models for surgical risk assessment of extracapsular extension (ECE) in patients with prostate cancer (PCa) before radical prostatectomy; and to compare the use of decision curve analysis (DCA) and receiver operating characteristic (ROC) metrics for selecting input feature combinations in models. METHODS: This retrospective observational study included two independent data sets: 139 participants from a single institution (training), and 55 from 15 other institutions (external validation), both treated with Robotic Assisted Radical Prostatectomy (RARP). Five ML models, based on different combinations of clinical, semantic (interpreted by a radiologist) and radiomics features computed from T2W-MRI images, were built to predict extracapsular extension in the prostatectomy specimen (pECE+). DCA plots were used to rank the models' net benefit when assigning patients to prostatectomy with non-nerve-sparing surgery (NNSS) or nerve-sparing surgery (NSS), depending on the predicted ECE status. DCA model rankings were compared with those drived from ROC area under the curve (AUC). RESULTS: In the training data, the model using clinical, semantic, and radiomics features gave the highest net benefit values across relevant threshold probabilities, and similar decision curve was observed in the external validation data. The model ranking using the AUC was different in the discovery group and favoured the model using clinical + semantic features only. CONCLUSIONS: The combined model based on clinical, semantic and radiomic features may be used to predict pECE + in patients with PCa and results in a positive net benefit when used to choose between prostatectomy with NNS or NNSS.

Prediction of extracapsular extension of prostate cancer by MRI radiomic signature: a systematic review.

The objective of this review is to survey radiomics signatures for detecting pathological extracapsular extension (pECE) on magnetic resonance imaging (MRI) in patients with prostate cancer (PCa) who underwent prostatectomy. Scientific Literature databases were used to search studies published from January 2007 to October 2023. All studies related to PCa MRI staging and using radiomics signatures to detect pECE after prostatectomy were included. Systematic review was performed according to Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA). The risk of bias and certainty of the evidence was assessed using QUADAS-2 and the radiomics quality score. From 1247 article titles screened, 16 reports were assessed for eligibility, and 11 studies were included in this systematic review. All used a retrospective study design and most of them used 3 T MRI. Only two studies were performed in more than one institution. The highest AUC of a model using only radiomics features was 0.85, for the test validation. The AUC for best model performance (radiomics associated with clinical/semantic features) varied from 0.72-0.92 and 0.69-0.89 for the training and validation group, respectively. Combined models performed better than radiomics signatures alone for detecting ECE. Most of the studies showed a low to medium risk of bias. After thorough analysis, we found no strong evidence supporting the clinical use of radiomics signatures for identifying extracapsular extension (ECE) in pre-surgery PCa patients. Future studies should adopt prospective multicentre approaches using large public datasets and combined models for detecting ECE. CRITICAL RELEVANT STATEMENT: The use of radiomics algorithms, with clinical and AI integration, in predicting extracapsular extension, could lead to the development of more accurate predictive models, which could help improve surgical planning and lead to better outcomes for prostate cancer patients. PROTOCOL OF SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42021272088. Published: https://doi.org/10.1136/bmjopen-2021-052342 . KEY POINTS: Radiomics can extract diagnostic features from MRI to enhance prostate cancer diagnosis performance. The combined models performed better than radiomics signatures alone for detecting extracapsular extension. Radiomics are not yet reliable for extracapsular detection in PCa patients.

MRI Assessed Placental Location as a Diagnostic Tool of Placental Invasiveness and Maternal Peripartum Morbidity.

Placenta accreta spectrum (PAS) disorder is one of the leading causes of peripartum maternal morbidity and mortality; its early identification during pregnancy is of utmost importance to ensure the optimal clinical outcome. The aim of the present study is to investigate the possible association of the presence and type/location of placenta previa on MRI with PAS and maternal peripartum outcome. One hundred eighty-nine pregnant women (mean age: 35 years; mean gestational age: 32 weeks) at high risk for PAS underwent a dedicated placental MRI. All women underwent a C-section within 6 weeks from the MRI. All MRIs were evaluated by two experienced genitourinary radiologists for presence, type (complete/partial vs. marginal/low lying), and location (anterior vs. anterior-posterior vs. posterior) of placenta previa. Statistical analysis was performed for possible association of type/location of previa with placental invasiveness and peripartum outcomes. Intraoperative information was used as a reference standard. Complete/partial previa was detected in 143/189 (75.6%) and marginal/low lying previa in 33/189 (17.5%) women; in 88/189 (46.6%) women, the placenta had anterior-posterior, in 54/189 (28.6%) anterior and in 41/189 (21.7%) posterior. Complete/partial previa had an at least 3-fold probability of invasiveness and was more frequently associated with unfavorable peripartum events, including massive intraoperative blood loss or hysterectomy, compared to low-lying/marginal placenta. Posterior placental location was significantly associated with lower rates of PAS and better clinical outcomes. In conclusion, the type and location of placenta previa shown with MRI seems to be associated with severity of complications during delivery and should be carefully studied.

Manifestations of Ovarian Cancer in Relation to Other Pelvic Diseases by MRI.

Imaging plays a pivotal role in the diagnostic approach of women with suspected ovarian cancer. MRI is widely used for preoperative characterization and risk stratification of adnexal masses. While epithelial ovarian cancer (EOC) has typical findings on MRI; there are several benign and malignant pelvic conditions that may mimic its appearance on imaging. Knowledge of the origin and imaging characteristics of a pelvic mass will help radiologists diagnose ovarian cancer promptly and accurately. Finally, in special subgroups, including adolescents and gravid population, the prevalence of various ovarian tumors differs from that of the general population and there are conditions which uniquely manifest during these periods of life.

Hepatokine Profile in Adolescents with Polycystic Ovary Syndrome: A Case-Control Study.

The current guidelines suggest routine screening for non-alcoholic fatty liver disease (NAFLD) in patients with polycystic ovary syndrome (PCOS). Hepatokines seem to be promising surrogate endpoints for the diagnosis and severity of NAFLD. PCOS has its onset in adolescence and its metabolic sequalae begin during the same period. There are scarce data on the hepatokine profile of adolescent PCOS patients. This case-control study examined the serum profile of the hepatokines sex hormone-binding globulin (SHBG), selenoprotein P, fibroblast growth factor 21 (FGF21), and fetuin A in a sample of adolescent PCOS patients, and their association to metabolic and hormonal parameters. The selenoprotein P and SHBG serum concentrations were significantly decreased in PCOS patients vs. the controls (median (IQR), 2.47 (0.40) vs. 2.66 (0.36) µg/mL, p = 0.025; mean ± SD, 41.71 ± 19.41 vs. 54.94 ± 22.12 nmol/L, p = 0.011, respectively), whereas selenoprotein P was significantly and positively associated with testosterone (r = 0.325, p = 0.007) and the free androgen index (r = 0.361, p = 0.002). The SHBG demonstrated multiple significant negative correlations with adverse metabolic parameters. Among the PCOS patients, the FGF21 concentrations were significantly higher in those with NAFLD, whereas a 1 pg/mL increase in the FGF21 concentration increased the odds of NAFLD diagnosis by liver ultrasound by 1%, suggesting FGF21 as a potential biomarker for hepatic disease in females with PCOS in adolescence. Fetuin A was the least differentiated hepatokine between the PCOS patients and controls with the least associations with metabolic and hormonal parameters.

The Burden of Non-Alcoholic Fatty Liver Disease in Adolescents with Polycystic Ovary Syndrome: A Case-Control Study.

The aim of this case-control study was to assess the burden of non-alcoholic fatty liver disease (NAFLD) in adolescents with polycystic ovary syndrome (PCOS) and its associations with insulin resistance, hyperandrogenism, and other metabolic characteristics of the syndrome. A total of 87 Caucasian adolescent girls (47 with PCOS and 40 controls), aged 12.3-20.4 years, underwent blood sampling for glucose metabolism, hormonal and lipid profile, gynecological and liver ultrasound, and liver elastography. Indices of insulin resistance, liver steatosis, and liver fibrosis were calculated. NAFLD diagnosed by ultrasound was more prevalent in adolescents with PCOS than controls (22.7% vs. 6.1%, p = 0.046), and was also verified by liver steatosis indices. The latter was not apparent for hepatic fibrosis, as assessed by Fibroscan® and calculated indices. The homeostatic model assessment for insulin resistance (HOMA-IR) was found to predict NAFLD diagnosis by the liver fat score (LFS) index (ß = 0.709, p = 0.002). Adolescents with PCOS and high free androgen index (FAI) presented worse NAFLD than those adolescents with PCOS and lower FAI. In addition, adolescents with PCOS and concurrent NAFLD had worse insulin sensitivity indices (HOMA-IR, QUICKI, and glucose to insulin ratio) than adolescents with PCOS alone. Adolescent insulin resistance could be considered a confounder of the association between PCOS and NAFLD.

A Novel Approach for Estimating Ovarian Cancer Tissue Heterogeneity through the Application of Image Processing Techniques and Artificial Intelligence.

PURPOSE: Tumor heterogeneity may be responsible for poor response to treatment and adverse prognosis in women with HGOEC. The purpose of this study is to propose an automated classification system that allows medical experts to automatically identify intratumoral areas of different cellularity indicative of tumor heterogeneity. METHODS: Twenty-two patients underwent dedicated pelvic MRI, and a database of 11,095 images was created. After image processing techniques were applied to align and assess the cancerous regions, two specific imaging series were used to extract quantitative features (radiomics). These features were employed to create, through artificial intelligence, an estimator of the highly cellular intratumoral area as defined by arbitrarily selected apparent diffusion coefficient (ADC) cut-off values (ADC < 0.85 × 10-3 mm2/s). RESULTS: The average recorded accuracy of the proposed automated classification system was equal to 0.86. CONCLUSION: The proposed classification system for assessing highly cellular intratumoral areas, based on radiomics, may be used as a tool for assessing tumor heterogeneity.

Comparison between 1.5-T and 3.0-T MRI for the diagnosis of placenta accreta spectrum disorders.

PURPOSE: Accurate antenatal diagnosis of placenta accreta spectrum (PAS) is important for optimal management. The purpose of this study was to compare the respective capabilities of 1.5-T and 3.0-T MRI in the diagnosis of PAS. MATERIALS AND METHODS: Between March 2016-March 2021, 190 pregnant women at high risk for PAS underwent dedicated prenatal MRI with either 1.5-T or 3.0-T units at a tertiary imaging center. Cesarian section and MRI were performed less than 6 weeks from each other. Prospectively collected data were evaluated by two experienced genitourinary radiologists for presence and extent of PAS. A comparative study was designed to investigate differences in predictive ability between 1.5-T and 3.0-T MRI groups. Sensitivity, specificity, accuracy, negative and positive prognostic values relative to intraoperative/histological findings, were computed for both groups and were compared with chi-square (χ 2) test. Interobserver agreement was estimated using Kappa test. RESULTS: One hundred-eighty-two gravid women were included in the study; of these, 91/182 (50%) women were evaluated with 1.5-T (mean age, 35 ± 5.1 [SD] years; mean gestational age: 32.5 weeks) and 91/182 (50%) with 3.0-T MRI (mean age, 34.9 ± 4.9 [SD] years; mean gestational age, 32.1 weeks). 1.5-T MRI yielded 95.7% sensitivity (95% CI: 87.8-99.1) and 81.8% specificity (95% CI: 59.8) and 3.0-T MRI 93.8% sensitivity (95% CI: 86.0-97.9) and 83.3% specificity (95% CI: 48.2-97.7) for PAS identification, with no differences between the two groups (P = 0.725 and P >0.999, respectively). MRI showed excellent predictive ability for detecting extrauterine placental spread with 100% sensitivity (95% CI: 89.4-100.0), 96.7% specificity (95% CI: 88.1-99.6) for 1.5-T and 97% sensitivity (95% CI: 84.2-99.9), 96.7% specificity (95% CI: 88.1-99.6) for 3.0-T without differences between the two groups (P > 0.999). Interobserver agreement was excellent for both groups. The most frequently detected MRI signs of PAS for both 1.5-T and 3.0-T groups were placental heterogeneity (n = 85, 93.5% vs. n = 90, 98.9%; P = 0.413), and intraplacental fetal vessels (n = 64, 70.3% vs. n = 65, 71.4%; P = 0.870). CONCLUSION: This study suggests that 3.0-T MRI and 1.5-T MRI are equivalent for the diagnosis of PAS.

Imaging during pregnancy: What the radiologist needs to know.

During the last decades, there has been a growing demand for medical imaging in gravid women. Imaging of the pregnant woman is challenging as it involves both the mother and the fetus and, consequently, several medical, ethical, or legal considerations are likely to be raised. Theoretically, all currently available imaging modalities may be used for the evaluation of the pregnant woman; however, in practice, confusion regarding the safety of the fetus often results in unnecessary avoidance of useful diagnostic tests, especially those involving ionizing radiation. This review article is focused on the current safety guidelines and considerations regarding the use of different imaging modalities in the pregnant population; also presented is an imaging work-up for the most common medical conditions of pregnant women, with emphasis on fetal and maternal safety.

Imaging of Gynecologic Malignancy in a Reproductive Age Female: Cancer During Pregnancy.

Gynecologic malignancies are common among cancers diagnosed during pregnancy, especially those of cervical and ovarian origin. Imaging is an important part of the diagnosis, staging, and follow-up of pregnancy-associated gynecologic tumors, with sonography and magnetic resonance (MR) imaging being the most suitable modalities. MR imaging is particularly useful in cervical cancer for the evaluation of tumor size, nodal, and extrapelvic disease. Ovarian tumor is initially diagnosed with sonography; MR imaging should be performed in cases of indeterminate ultrasonography findings and for staging. Pregnancy-related changes may be responsible for erroneous diagnosis; radiologists should be aware of such pitfalls to avoid misinterpretation.