Comparison between Three Radiomics Models and Clinical Nomograms for Prediction of Lymph Node Involvement in PCa Patients Combining Clinical and Radiomic Features.

PURPOSE: We aim to compare the performance of three different radiomics models (logistic regression (LR), random forest (RF), and support vector machine (SVM)) and clinical nomograms (Briganti, MSKCC, Yale, and Roach) for predicting lymph node involvement (LNI) in prostate cancer (PCa) patients. MATERIALS AND METHODS: The retrospective study includes 95 patients who underwent mp-MRI and radical prostatectomy for PCa with pelvic lymphadenectomy. Imaging data (intensity in T2, DWI, ADC, and PIRADS), clinical data (age and pre-MRI PSA), histological data (Gleason score, TNM staging, histological type, capsule invasion, seminal vesicle invasion, and neurovascular bundle involvement), and clinical nomograms (Yale, Roach, MSKCC, and Briganti) were collected for each patient. Manual segmentation of the index lesions was performed for each patient using an open-source program (3D SLICER). Radiomic features were extracted for each segmentation using the Pyradiomics library for each sequence (T2, DWI, and ADC). The features were then selected and used to train and test three different radiomics models (LR, RF, and SVM) independently using ChatGPT software (v 4o). The coefficient value of each feature was calculated (significant value for coefficient ≥ ±0.5). The predictive performance of the radiomics models and clinical nomograms was assessed using accuracy and area under the curve (AUC) (significant value for p ≤ 0.05). Thus, the diagnostic accuracy between the radiomics and clinical models were compared. RESULTS: This study identified 343 features per patient (330 radiomics features and 13 clinical features). The most significant features were T2_nodulofirstordervariance and T2_nodulofirstorderkurtosis. The highest predictive performance was achieved by the RF model with DWI (accuracy 86%, AUC 0.89) and ADC (accuracy 89%, AUC 0.67). Clinical nomograms demonstrated satisfactory but lower predictive performance compared to the RF model in the DWI sequences. CONCLUSIONS: Among the prediction models developed using integrated data (radiomics and semantics), RF shows slightly higher diagnostic accuracy in terms of AUC compared to clinical nomograms in PCa lymph node involvement prediction.

Accuracy of magnetic resonance imaging to identify pseudocapsule invasion in renal tumors.

PURPOSE: To evaluate accuracy of MRI in detecting renal tumor pseudocapsule (PC) invasion and to propose a classification based on imaging of PC status in patients with renal cell carcinoma. METHODS: From January 2017 to June 2018, 58 consecutive patients with localized renal cell carcinoma were prospectively enrolled. MRI was performed preoperatively and PC was classified, according to its features, as follows: MRI-Cap 0 (absence of PC), MRI-Cap 1 (presence of a clearly identifiable PC), MRI-Cap 2 (focally interrupted PC), and MRI-Cap 3 (clearly interrupted and infiltrated PC). A 3D image reconstruction showing MRI-Cap score was provided to both surgeon and pathologist to obtain complete preoperative evaluation and to compare imaging and pathology reports. All patients underwent laparoscopic partial nephrectomy. In surgical specimens, PC was classified according to the renal tumor capsule invasion scoring system (i-Cap). RESULTS: A concordance between MRI-Cap and i-Cap was found in 50/58 (86%) cases. ρ coefficient for each MRI-cap and iCap categories was: MRI-Cap 0: 0.89 (p < 0.0001), MRI-Cap1: 0.75 (p < 0.0001), MRI-Cap 2: 0.76 (p < 0.0001), and MRI-Cap3: 0.87 (p < 0.0001). Sensitivity, specificity, positive predictive value, negative predictive value, and AUC were: MRI-Cap 0: Se 97.87% Spec 83.3%, PPV 95.8%, NPV 90.9%, and AUC 90.9; MRI-Cap 1: Se 77% Spec 95.5%, PPV 83.3%, NPV 93.5%, and AUC 0.86; MRI-Cap 2- iCap 2: Se 88% Spec 90%, PPV 79%, NPV 95%, and AUC 0.89; MRI-Cap 3: Se 94% Spec 95%, PPV 88%, NPV 97%, and AUC 0.94. CONCLUSIONS: MRI-Cap classification is accurate in evaluating renal tumor PC features. PC features can provide an imaging-guided landmark to figure out where a minimal margin could be preferable during nephron-sparing surgery.

Diffusion-weighted lesions after carotid artery stenting are associated with cognitive impairment.

The effect of carotid artery stenting (CAS) on cognitive function is still debated. Cerebral microembolism, detectable by post-procedural diffusion-weighted imaging (DWI) lesions, has been suggested to predispose to cognitive decline. Our study aimed at evaluating the effect of CAS on cognitive profile focusing on the potential role of cerebral microembolic lesions, taking into consideration the impact of factors potentially influencing cognitive status (demographic features, vascular risk profile, neuropsychological evaluation at baseline and magnetic resonance (MR) markers of brain structural damage). Thirty-seven patients with severe carotid artery stenosis were enrolled. Neurological assessment, neuropsychological evaluation and brain MR were performed the day before CAS (E0). Brain MR with DWI was repeated the day after CAS (E1), while neuropsychological evaluation was done after a 14-month median period (E2). Volumes of both white matter hyperintensities and whole brain were estimated at E0 on axial MR FLAIR and T1w-SE sequences, respectively. Unadjusted ANOVA analysis showed a significant CAS*DWI interaction for MMSE (F=7.154(32), p=.012). After adjusting for factors potentially influencing cognitive status CAS*DWI interaction was confirmed for MMSE (F=7.092(13), p=.020). Patients with DWI lesions showed a mean E2-E0 MMSE reduction of -3.1, while group without DWI lesions showed a mean E2-E0 MMSE of +1.1. Our study showed that peri-procedural brain microembolic load impacts negatively on cognitive functions, independently from the influence of patients-related variables.

Determination of stature from skeletal and skull measurements by CT scan evaluation.

The aim of this article is to find a correlation between height and femur/skull measurements through Computed Tomography (CT) scans and derive regression equations for total skeletal height estimation in the Caucasian population. We selected 200 Caucasian patients from March 2010 to July 2011 who had to perform a CT scan for cancer restaging. The mean age is 64.5 years. Both sexes are represented by the same number of persons. Patients have executed a total body CT scan with contrast; once scan accomplished, we measured height through a digital scales. We analyzed CT scans of each patient, obtaining multiplanar reconstruction in sagittal and coronal planes with 1mm of thickness, and we measured 10 diameters of skull and femur. Then we performed a single and a multiple regression analysis considering the three diameters that better correlated with height. The skeletal diameters with the highest correlation coefficients with stature were femur lengths, length of cranial base (Ba-N), and distance from the posterior extremity of the cranial base to the inferior point of the nasal bone (Ba-NB). Although both femur and skull are skeletal segments used for stature estimation, in our sample femur gave stronger correlation with height than skull. h=35.7+1.48·BaN+2.32·BaNB+2.53·FEM and h=3.06·FEM+72.6 are the formulae that provided the most accurate stature assessment using multiple and single regression analysis respectively.