作者: Hassan M. Fathallah-Shaykh , Andrew DeAtkine , Elizabeth Coffee , Elias Khayat , Asim K. Bag
DOI: 10.1371/JOURNAL.PMED.1002810
关键词: Radiation therapy 、 Glioma 、 Gold standard (test) 、 Predictive value of tests 、 Quality of life 、 Medicine 、 Stage (cooking) 、 Radiology 、 Magnetic resonance imaging 、 Retrospective cohort study
摘要: Background Low-grade gliomas cause significant neurological morbidity by brain invasion. There is no universally accepted objective technique available for detection of enlargement low-grade in the clinical setting; subjective evaluation clinicians using visual comparison longitudinal radiological studies gold standard. The aim this study to determine whether a computer-assisted diagnosis (CAD) method helps physicians detect earlier growth gliomas. Methods and findings We reviewed 165 patients diagnosed with grade 2 gliomas, seen at University Alabama Birmingham clinics from 1 July 2017 14 May 2018. MRI scans were collected during spring summer Fifty-six met inclusion criteria, including 19 oligodendrogliomas, 26 astrocytomas, 11 mixed 30 males females mean age 48 years range follow-up 150.2 months (difference between highest lowest values). None received radiation therapy. also studied 7 an imaging abnormality without pathological diagnosis, who clinically stable time retrospective review (14 2018). This compared aided CAD reports. tumors 63 (56 + 7) 627 digitized, 34 progression 22 radiologically consisted tumor segmentation, computing volumes, pointing online abrupt change-of-point method, which considers only past measurements. Independent scientists have evaluated segmentation method. In 29 progression, median was 44 current standard care (p < 0.001). Using CAD, accurate possible 57% change volume as 174% necessary diagnose methods group, facilitated 13 out patients. did not group. main limitation its design; nevertheless, results depict state practice that allowed increase volumes baseline before detection. Such large increases would be permitted prospective design. number glioma (n = 56) limitation; however, it equivalent phase II trials. Conclusions associated delays detecting Our support idea significantly smaller than alone. does answer questions treat or treatment modality optimal. Nonetheless, early sets stage future address these therapeutic interventions prolong survival improve quality life.