Deep learning assistance increases the detection sensitivity of radiologists for secondary intracranial aneurysms in subarachnoid hemorrhage.

PURPOSE: To evaluate whether a deep learning model (DLM) could increase the detection sensitivity of radiologists for intracranial aneurysms on CT angiography (CTA) in aneurysmal subarachnoid hemorrhage (aSAH). METHODS: Three different DLMs were trained on CTA datasets of 68 aSAH patients with 79 aneurysms with their outputs being combined applying ensemble learning (DLM-Ens). The DLM-Ens was evaluated on an independent test set of 104 aSAH patients with 126 aneuryms (mean volume 129.2 ± 185.4 mm3, 13.0% at the posterior circulation), which were determined by two radiologists and one neurosurgeon in consensus using CTA and digital subtraction angiography scans. CTA scans of the test set were then presented to three blinded radiologists (reader 1: 13, reader 2: 4, and reader 3: 3 years of experience in diagnostic neuroradiology), who assessed them individually for aneurysms. Detection sensitivities for aneurysms of the readers with and without the assistance of the DLM were compared. RESULTS: In the test set, the detection sensitivity of the DLM-Ens (85.7%) was comparable to the radiologists (reader 1: 91.2%, reader 2: 86.5%, and reader 3: 86.5%; Fleiss κ of 0.502). DLM-assistance significantly increased the detection sensitivity (reader 1: 97.6%, reader 2: 97.6%,and reader 3: 96.0%; overall P=.024; Fleiss κ of 0.878), especially for secondary aneurysms (88.2% of the additional aneurysms provided by the DLM). CONCLUSION: Deep learning significantly improved the detection sensitivity of radiologists for aneurysms in aSAH, especially for secondary aneurysms. It therefore represents a valuable adjunct for physicians to establish an accurate diagnosis in order to optimize patient treatment.

Predicting SGA neonates using first-trimester screening: influence of previous pregnancy's birthweight and PAPP-A MoM.

OBJECTIVE: Investigating the proportions of anamnestic and biochemical variables of the previous and current pregnancies for the prediction of small for gestational age (SGA) neonates in the current pregnancy. METHODS: In this observational retrospective study, 45 029 pregnancies were examined, including 3862 patients with more than one pregnancy. Odds ratios for SGA using anamnestic parameters and pregnancy-associated plasma protein A (PAPP-A) values from all pregnancies were estimated by using a logistic regression model. RESULTS: There were 2552 (5.7%) SGA neonates. Two threshold PAPP-A values were identified at 0.15 MoM and 0.33 MoM with probabilities for SGA of 23% and 17%, respectively. A previous SGA < 10th centile and a current PAPP-A MoM value < 5th centile result in odds ratios of 4.8 (95% CI: 3.5-6.5) and 3.0 (95% CI: 1.8-5.0), respectively. The parameters' combined odds ratio is 14.1 (95% CI: 3.9-50.3) with a number needed to screen of ten for one SGA neonate at a detection rate of 37%. CONCLUSION: Information on previous pregnancies affected by SGA and a current pregnancy's low PAPP-A value are reliable predictors for a SGA delivery. First-trimester biochemical analysis should be maintained to detect women at risk for delivering a SGA neonate.

The mitochondrial permeability transition from in vitro artifact to disease target.

The mitochondrial permeability transition pore is a high conductance channel whose opening leads to an increase of mitochondrial inner membrane permeability to solutes with molecular masses up to approximately 1500 Da. In this review we trace the rise of the permeability transition pore from the status of in vitro artifact to that of effector mechanism of cell death. We then cover recent results based on genetic inactivation of putative permeability transition pore components, and discuss their meaning for our understanding of pore structure. Finally, we discuss evidence indicating that the permeability transition pore plays a role in pathophysiology, with specific emphasis on in vivo models of disease.

Properties of the permeability transition in VDAC1(-/-) mitochondria.

Opening of the permeability transition pore (PTP), a high-conductance mitochondrial channel, causes mitochondrial dysfunction with Ca2+ deregulation, ATP depletion, release of pyridine nucleotides and of mitochondrial apoptogenic proteins. Despite major efforts, the molecular nature of the PTP remains elusive. A compound library screening led to the identification of a novel high affinity PTP inhibitor (Ro 68-3400), which labeled a approximately 32 kDa protein that was identified as isoform 1 of the voltage-dependent anion channel (VDAC1) [A.M. Cesura, E. Pinard, R. Schubenel, V. Goetschy, A. Friedlein, H. Langen, P. Polcic, M.A. Forte, P. Bernardi, J.A. Kemp, The voltage-dependent anion channel is the target for a new class of inhibitors of the mitochondrial permeability transition pore. J. Biol. Chem. 278 (2003) 49812-49818]. In order to assess the role of VDAC1 in PTP formation and activity, we have studied the properties of mitochondria from VDAC1(-/-) mice. The basic properties of the PTP in VDAC1(-/-) mitochondria were indistinguishable from those of strain-matched mitochondria from wild-type CD1 mice, including inhibition by Ro 68-3400, which labeled identical proteins of 32 kDa in both wild-type and VDAC1(-/-) mitochondria. The labeled protein could be separated from all VDAC isoforms. While these results do not allow to exclude that VDAC is part of the PTP, they suggest that VDAC is not the target for PTP inhibition by Ro 68-3400.

Cyclosporin A generates superoxide in smooth muscle cells.

Cyclosporin A (CsA) generates superoxide in smooth muscle cells. Our earlier studies have demonstrated that the increase in the vasopressin type 1 receptor induced in vascular smooth muscle cells in the presence of CsA is probably due to superoxide (Krauskopf et al., J Biol Chem 278, 41685-41690, 2003). This increase in vasopressin receptor is likely at the base of increased vascular responsiveness to vasoconstrictor hormones and hypertension induced by CsA. Here, we demonstrate that CsA produces superoxide. In addition, our data show that superoxide generation does not originate from the major cellular superoxide generating systems NAD(P)H oxidase or xanthine oxidase. Our results suggest that the side effects of CsA could be diminished with the help of SOD mimetic drugs.

Role of superoxide as a signaling molecule.

Superoxide is known to affect vascular physiology in several ways and has also been recognized to contribute significantly to vascular physiopathology. Here we discuss the emerging role of superoxide as an essential signaling molecule in normal physiology.

Vasopressin type 1A receptor up-regulation by cyclosporin A in vascular smooth muscle cells is mediated by superoxide.

Based on our previous results, we investigated whether cyclosporin A (CsA)-induced vasopressin type 1A receptor up-regulation was mediated by free radicals. We report that CsA analogues with different affinities for cyclophilin and calcineurin were able to up-regulate vasopressin type 1A receptor and to generate free radicals in smooth muscle cells independently of calcineurin. Further, we demonstrate that the antioxidant N-acetyl-L-cysteine blocked the increase in vasopressin type 1A receptor mRNA and protein levels induced by CsA and that low concentrations of prooxidants were able to directly increase vasopressin type 1A receptor mRNA and protein levels. In addition, short exposure to CsA or pro-oxidants was sufficient to significantly increase vasopressin type 1A receptor mRNA and protein levels. Using cell-permeable forms of superoxide dismutase and catalase, we finally show that superoxide mediates the CsA-induced effects on vasopressin type 1A receptor. These results provide strong evidence that CsA-induced superoxide generation is causally involved in vasopressin type 1A receptor expression and demonstrate for the first time that low physiological concentrations of radicals, most probably superoxide, are able to directly affect cellular signaling to increase vasopressin type 1A receptor expression in rat aortic smooth muscle cells.

Cyclosporin A-induced free radical generation is not mediated by cytochrome P-450.

1. Reactive oxygen species (ROS) have been proposed to play a role in the side effects of the immunosuppressive drug cyclosporin A (CsA). 2. The aim of this study was to investigate whether cytochrome P-450 (CYP) dependent metabolism of CsA could be responsible for ROS generation since it has been suggested that CsA may influence the CYP system to produce ROS. 3. We show that CsA (1 -- 10 microM) generated antioxidant-inhibitable ROS in rat aortic smooth muscle cells (RASMC) using the fluorescent probe 2,7-dichlorofluorescin diacetate. 4. Using cytochrome c as substrate, we show that CsA (10 microM) did not inhibit NADPH cytochrome P-450 reductase in microsomes prepared from rat liver, kidney or RASMC. 5. CsA (10 microM) did not uncouple the electron flow from NADPH via NADPH cytochrome P-450 reductase to the CYP enzymes because CsA did not inhibit the metabolism of substrates selective for several CYP enzymes that do not metabolize CsA in rat liver microsomes. 6. CsA (10 microM) did not generate more radicals in CYP 3A4 expressing immortalized human liver epithelial cells (T5-3A4 cells) than in control cells that do not express CYP 3A4. 7. Neither diphenylene iodonium nor the CYP 3A inhibitor ketoconazole were able to block ROS formation in rat aortic smooth muscle or T5-3A4 cells. 8. These results demonstrate that CYP enzymes do not contribute to CsA-induced ROS formation and that CsA neither inhibits NADPH cytochrome P-450 reductase nor the electron transfer to the CYP enzymes.