Value of [18F]FDG PET-CT in the follow-up of surgically treated oral tongue squamous cell carcinoma: single centre cohort analysis on 87 patients.

BACKGROUND: To evaluate the diagnostic performance of [¹8F]fluorodeoxyglucose positron emission tomography/computed tomography ([¹8F]FDG-PET/CT) scan in detecting local recurrences in patients with surgically treated oral tongue squamous cell cancer (OTSCC). MATERIAL AND METHODS: Eighty-seven patients who had undergone surgery for OTSCC were monitored clinically and [¹8F]FDGPET/CT and magnetic resonance (MR). PET uptakes were classified as functional (Type A), suspicious (Type B), or highly suggestive of local recurrence (Type C). A multidisciplinary team (MDT) evaluated case-by-case the surveillance strategy based on PET uptake. RESULTS: Fifty-nine patients presented FDG-PET uptake during follow-up: this report was significantly more frequent in patients who received flap reconstruction than in those without (73% vs 50%; p = 0.05). In 13 patients with Type A (n = 1), Type B (n = 9), and Type C (n = 3) uptakes an additional MR was considered preferable and discovered recurrence in 12.PET-CT had 9 true positives, 17 false positives, 71 true negatives, and no false-negative, resulting in sensitivity, specificity, positive (PPV) and negative predictive values (NPV) of 100%, 80.7%, 34.6%, and 100%. CONCLUSIONS: The present results demonstrated a change in diagnostic strategy, as decided by the MDT, in about one-fifth of patients. The results should prompt in designing a rational surveillance schedule in surgically treated OTSCC.

Automatic classification of radiological reports for clinical care.

Radiological reporting generates a large amount of free-text clinical narratives, a potentially valuable source of information for improving clinical care and supporting research. The use of automatic techniques to analyze such reports is necessary to make their content effectively available to radiologists in an aggregated form. In this paper we focus on the classification of chest computed tomography reports according to a classification schema proposed for this task by radiologists of the Italian hospital ASST Spedali Civili di Brescia. The proposed system is built exploiting a training data set containing reports annotated by radiologists. Each report is classified according to the schema developed by radiologists and textual evidences are marked in the report. The annotations are then used to train different machine learning based classifiers. We present in this paper a method based on a cascade of classifiers which make use of a set of syntactic and semantic features. The resulting system is a novel hierarchical classification system for the given task, that we have experimentally evaluated.