Valeria Romeo

Bio: Valeria Romeo is an academic researcher from University of Naples Federico II. The author has contributed to research in topics: Medicine & Magnetic resonance imaging. The author has an hindex of 15, co-authored 85 publications receiving 705 citations. Previous affiliations of Valeria Romeo include Medical University of Vienna.

Machine learning applications in prostate cancer magnetic resonance imaging

Abstract: With this review, we aimed to provide a synopsis of recently proposed applications of machine learning (ML) in radiology focusing on prostate magnetic resonance imaging (MRI). After defining the difference between ML and classical rule-based algorithms and the distinction among supervised, unsupervised and reinforcement learning, we explain the characteristic of deep learning (DL), a particular new type of ML, including its structure mimicking human neural networks and its ‘black box’ nature. Differences in the pipeline for applying ML and DL to prostate MRI are highlighted. The following potential clinical applications in different settings are outlined, many of them based only on MRI-unenhanced sequences: gland segmentation; assessment of lesion aggressiveness to distinguish between clinically significant and indolent cancers, allowing for active surveillance; cancer detection/diagnosis and localisation (transition versus peripheral zone, use of prostate imaging reporting and data system (PI-RADS) version 2), reading reproducibility, differentiation of cancers from prostatitis benign hyperplasia; local staging and pre-treatment assessment (detection of extraprostatic disease extension, planning of radiation therapy); and prediction of biochemical recurrence. Results are promising, but clinical applicability still requires more robust validation across scanner vendors, field strengths and institutions.

Prediction of Tumor Grade and Nodal Status in Oropharyngeal and Oral Cavity Squamous-cell Carcinoma Using a Radiomic Approach.

Abstract: BACKGROUND/AIM To investigate whether a radiomic machine learning (ML) approach employing texture-analysis (TA) features extracted from primary tumor lesions (PTLs) is able to predict tumor grade (TG) and nodal status (NS) in patients with oropharyngeal (OP) and oral cavity (OC) squamous-cell carcinoma (SCC). PATIENTS AND METHODS Contrast-enhanced CT images of 40 patients with OP and OC SCC were post-processed to extract TA features from PTLs. A feature selection method and different ML algorithms were applied to find the most accurate subset of features to predict TG and NS. RESULTS For the prediction of TG, the best accuracy (92.9%) was achieved by Naive Bayes (NB), bagging of NB and K Nearest Neighbor (KNN). For the prediction of NS, J48, NB, bagging of NB and boosting of J48 overcame the accuracy of 90%. CONCLUSION A radiomic ML approach applied to PTLs is able to predict TG and NS in patients with OC and OP SCC.

European Society of Endocrinology Clinical Practice Guidelines on the management of adrenocortical carcinoma in adults, in collaboration with the European Network for the Study of Adrenal Tumors

Abstract: Adrenocortical carcinoma (ACC) is a rare and in most cases steroid hormone-producing tumor with variable prognosis. The purpose of these guidelines is to provide clinicians with best possible evidence-based recommendations for clinical management of patients with ACC based on the GRADE (Grading of Recommendations Assessment, Development and Evaluation) system. We predefined four main clinical questions, which we judged as particularly important for the management of ACC patients and performed systematic literature searches: (A) What is needed to diagnose an ACC by histopathology? (B) Which are the best prognostic markers in ACC? (C) Is adjuvant therapy able to prevent recurrent disease or reduce mortality after radical resection? (D) What is the best treatment option for macroscopically incompletely resected, recurrent or metastatic disease? Other relevant questions were discussed within the group. Selected Recommendations: (i) We recommend that all patients with suspected and proven ACC are discussed in a multidisciplinary expert team meeting. (ii) We recommend that every patient with (suspected) ACC should undergo careful clinical assessment, detailed endocrine work-up to identify autonomous hormone excess and adrenal-focused imaging. (iii) We recommend that adrenal surgery for (suspected) ACC should be performed only by surgeons experienced in adrenal and oncological surgery aiming at a complete en bloc resection (including resection of oligo-metastatic disease). (iv) We suggest that all suspected ACC should be reviewed by an expert adrenal pathologist using the Weiss score and providing Ki67 index. (v) We suggest adjuvant mitotane treatment in patients after radical surgery that have a perceived high risk of recurrence (ENSAT stage III, or R1 resection, or Ki67 >10%). (vi) For advanced ACC not amenable to complete surgical resection, local therapeutic measures (e.g. radiation therapy, radiofrequency ablation, chemoembolization) are of particular value. However, we suggest against the routine use of adrenal surgery in case of widespread metastatic disease. In these patients, we recommend either mitotane monotherapy or mitotane, etoposide, doxorubicin and cisplatin depending on prognostic parameters. In selected patients with a good response, surgery may be subsequently considered. (vii) In patients with recurrent disease and a disease-free interval of at least 12 months, in whom a complete resection/ablation seems feasible, we recommend surgery or alternatively other local therapies. Furthermore, we offer detailed recommendations about the management of mitotane treatment and other supportive therapies. Finally, we suggest directions for future research.

Mural inflammation in Crohn’s disease: Location matched histological validation of MR imaging features

Abstract: PURPOSE To validate proposed magnetic resonance (MR) imaging features of Crohn disease activity against a histopathologic reference. MATERIALS AND METHODS Ethical permission was given by the University College London hospital ethics committee, and informed written consent was obtained from all participants. Preoperative MR imaging was performed in 18 consecutive patients with Crohn disease undergoing elective small-bowel resection. The Harvey-Bradshaw index, the C-reactive protein level, and disease chronicity were recorded. The resected bowel was retrospectively identified at preoperative MR imaging, and wall thickness, mural and lymph node/cerebrospinal fluid (CSF) signal intensity ratios on T2-weighted fat-saturated images, gadolinium-based contrast material uptake, enhancement pattern, and mesenteric signal intensity on T2-weighted fat-saturated images were recorded. Precise histologic matching was achieved by imaging the ex vivo surgical specimens. Histopathologic grading of acute inflammation with the acute inflammatory score (AIS) (on the basis of mucosal ulceration, edema, and quantity and depth of neutrophilic infiltration) and the degree of fibrostenosis was performed at each site, and results were compared with MR imaging features. Data were analyzed by using linear regression with robust standard errors of the estimate. RESULTS AIS was positively correlated with mural thickness and mural/CSF signal intensity ratio on T2-weighted fat-saturated images (P < .001 and P = .003, respectively) but not with mural enhancement at 30 and 70 seconds (P = .50 and P = .73, respectively). AIS was higher with layered mural enhancement (P < .001), a pattern also commonly associated with coexisting fibrostenosis (75%). Mural/CSF signal intensity ratio on T2-weighted fat-saturated images was higher in histologically edematous bowel than in nonedematous bowel (P = .04). There was no correlation between any lymph node characteristic and AIS. CONCLUSION Increasing mural thickness, high mural signal intensity on T2-weighted fat-saturated images, and a layered pattern of enhancement reflect histologic features of acute small-bowel inflammation in Crohn disease.

Drug Resistance in Glioblastoma: A Mini Review

Abstract: Glioblastoma multiforme (GBM) is recognized as the most common and lethal form of central nervous system cancer. Currently used surgical techniques, chemotherapeutic agents, and radiotherapy strategies have done very little in extending the life expectancies of patients diagnosed with GBM. The difficulty in treating this malignant disease lies both in its inherent complexity and numerous mechanisms of drug resistance. In this review, we summarize several of the primary mechanisms of drug resistance. We reviewed available published literature in the English language regarding drug resistance in glioblastoma. The reasons for drug resistance in glioblastoma include drug efflux, hypoxic areas of tumor cells, cancer stem cells, DNA damage repair, and miRNAs. Many potential therapies target these mechanisms, including a series of investigated alternative and plant-derived agents. Future research and clinical trials in glioblastoma patients should pursue combination of therapies to help combat drug resistance. The emerging new data on the potential of plant-derived therapeutics should also be closely considered and further investigated.