Deep learning LI-RADS grading system based on contrast enhanced multiphase MRI for differentiation between LR-3 and LR-4/LR-5 liver tumors.

BACKGROUND: To develop a deep learning (DL) method based on multiphase, contrast-enhanced (CE) magnetic resonance imaging (MRI) to distinguish Liver Imaging Reporting and Data System (LI-RADS) grade 3 (LR-3) liver tumors from combined higher-grades 4 and 5 (LR-4/LR-5) tumors for hepatocellular carcinoma (HCC) diagnosis. METHODS: A total of 89 untreated LI-RADS-graded liver tumors (35 LR-3, 14 LR-4, and 40 LR-5) were identified based on the radiology MRI interpretation reports. Multiphase 3D T1-weighted gradient echo imaging was acquired at six time points: pre-contrast, four phases immediately post-contrast, and one hepatobiliary phase after intravenous injection of gadoxetate disodium. Image co-registration was performed across all phases on the center tumor slice to correct motion. A rectangular tumor box centered on the tumor area was drawn to extract subset tumor images for each imaging phase, which were used as the inputs to a convolutional neural network (CNN). The pre-trained AlexNet CNN model underwent transfer learning using liver MRI data for LI-RADS tumor grade classification. The output probability number closer to 1 or 0 indicated a higher possibility of being combined LR-4/LR-5 tumor or LR-3 tumor, respectively. Five-fold cross validation was used for training (60% dataset), validation (20%) and testing processes (20%). RESULTS: The DL CNN model for LI-RADS grading using inputs of multiphase liver MRI data acquired at three time points (pre-contrast, arterial, and washout phase) achieved a high accuracy of 0.90, sensitivity of 1.0, precision of 0.835, and AUC of 0.95 with reference to the expert human radiologist report. The CNN output of probability provided radiologists a confidence level of the model's grading for each liver lesion. CONCLUSIONS: An AlexNet CNN model for LI-RADS grading of liver lesions provided diagnostic performance comparable to radiologists and offered valuable clinical guidance for differentiating intermediate LR-3 liver lesions from more-likely malignant LR-4/LR-5 lesions in HCC diagnosis.

Should 68Ga-DOTATATE PET/CT be Performed Routinely in Patients with Neuroendocrine Tumors Before Surgical Resection?

BACKGROUND: The only potential cure for neuroendocrine tumors (NETs) is operative resection, which may also offer a survival benefit for advanced disease. We aimed to assess the role of 68Ga-DOTATATE PET/CT in preoperative planning and compared its performance to CT with IV contrast and MRI with Eovist®, for abdominal NETs. METHODS: Records of patients who underwent 68Ga-DOTATATE PET/CT in addition to MRI with Eovist® and/or CT with IV contrast were retrospectively evaluated. The effect of imaging findings on surgical management and characteristics of detected lesions were analyzed. Descriptive statistics were used. RESULTS: Of 21 patients who underwent 68Ga-DOTATATE PET/CT prior to surgical resection, five (24%) had a change in surgical management due to findings. In three patients, 68Ga-DOTATATE PET/CT identified the primary tumor. In two patients, 68Ga-DOTATATE PET/CT helped clarify equivocal hepatic lesions seen on MRI with Eovist®. MRI with Eovist® had the highest number of lesions found (median 13, versus 9 on CT and 9.5 on 68Ga-DOTATATE PET/CT). DOTATATE-avid lesions were on average larger than lesions seen only on MRI with Eovist® (1.6 cm versus 0.6 cm, p = 0.0002). The optimal cutoff point for detection by 68Ga-DOTATATE PET/CT was a size of 0.95 cm, with a sensitivity of 56% and specificity of 98%. CONCLUSIONS: Preoperative 68Ga-DOTATATE PET/CT is useful only in a subset of patients undergoing surgical resection for NETs. MRI with Eovist® is superior at identifying liver metastases when compared to 68Ga-DOTATATE PET/CT and should therefore be used routinely before hepatic cytoreduction of NETs.

Regulation of the cellular heat shock response in Caenorhabditis elegans by thermosensory neurons.

Temperature pervasively affects all cellular processes. In response to a rapid increase in temperature, all cells undergo a heat shock response, an ancient and highly conserved program of stress-inducible gene expression, to reestablish cellular homeostasis. In isolated cells, the heat shock response is initiated by the presence of misfolded proteins and therefore thought to be cell-autonomous. In contrast, we show that within the metazoan Caenorhabditis elegans, the heat shock response of somatic cells is not cell-autonomous but rather depends on the thermosensory neuron, AFD, which senses ambient temperature and regulates temperature-dependent behavior. We propose a model whereby this loss of cell autonomy serves to integrate behavioral, metabolic, and stress-related responses to establish an organismal response to environmental change.