Deep learning of renal scans in children with antenatal hydronephrosis.

INTRODUCTION: Antenatal hydronephrosis (ANH) is one of the most common anomalies identified on prenatal ultrasound, found in up to 4.5% of all pregnancies. Children with ANH are surveilled with repeated renal ultrasound and when there is high suspicion for a ureteropelvic junction obstruction on renal ultrasound, a mercaptuacetyltriglycerine (MAG3) Lasix renal scan is performed to evaluate for obstruction. However, the challenging interpretation of MAG3 renal scans places patients at risk of misdiagnosis. OBJECTIVE: Our objective was to analyze MAG3 renal scans using machine learning to predict renal complications. We hypothesized that our deep learning model would extract features from MAG3 renal scans that can predict renal complications in children with ANH. STUDY DESIGN: We performed a case-control study of MAG3 studies drawn from a population of children with ANH concerning for ureteropelvic junction obstruction evaluated at our institution from January 2009 until June of 2021. The outcome was renal complications that occur ≥6 months after an equivocal MAG-3 renal scan. We created two machine learning models: a deep learning model using the radiotracer concentration versus time data from the kidney of interest and a random forest model created using clinical data. The performance of the models was assessed using measures of diagnostic accuracy. RESULTS: We identified 152 eligible patients with available images of which 62 were cases and 90 were controls. The deep learning model predicted future renal complications with an overall accuracy of 73% (95% confidence inteveral [CI] 68-76%) and an AUC of 0.78 (95% CI 0.7, 0.84). The random forest model had an accuracy of 62% (95% CI 60-66%) and an AUC of 0.67 (95% CI. 0 64, 0.72) DISCUSSION: Our deep learning model predicted patients at high risk of developing renal complications following an equivocal renal scan and discriminate those at low risk with moderately high accuracy (73%). The deep learning model outperformed the clinical model built from clinical features classically used by urologists for surgical decision making. CONCLUSION: Our models have the potential to influence clinical decision making by providing supplemental analytical data from MAG3 scans that would not otherwise be available to urologists. Future multi-institutional retrospective and prospective trials are needed to validate our model.

One-pot synthesis of nanochain particles for targeting brain tumors.

To synthesize multi-component nanochains, we developed a simple 'one-pot' synthesis, which exhibited high yield and consistency. The nanochains particles consist of parent nanospheres chemically linked into a higher-order, chain-like assembly. The one-pot synthesis is based on the addition of two types of parent nanospheres in terms of their surface chemical functionality (e.g., decorated with PEG-NH2 or PEG-COOH). By reacting the two types of parent nanospheres at a specific ratio (∼2 : 1) for a short period of time (∼30 min) under rigorous stirring, nanochains were formed. For example, we show the synthesis of iron oxide nanochains with lengths of about 125 nm consisting of 3-5 constituting nanospheres. The chain-like shaped nanoparticle possessed a unique ability to target and rapidly deposit on the endothelium of glioma sites via vascular targeting. To target and image invasive brain tumors, we used iron oxide nanochains with the targeting ligand being the fibronectin-targeting peptide CREKA. Overexpression of fibronectin is strongly associated with the perivascular regions of glioblastoma multiforme and plays a critical role in migrating and invasive glioma cells. In mice with invasive glioma tumors, 3.7% of the injected CREKA-targeted nanochains was found in gliomas within 1 h. Notably, the intratumoral deposition of the nanochain was ∼2.6-fold higher than its spherical variant. Using MR imaging, the precise targeting of nanochains to gliomas provided images with the exact topology of the disease including their margin of infiltrating edges and distant invasive sites.

CEST-FISP: a novel technique for rapid chemical exchange saturation transfer MRI at 7 T.

Chemical exchange saturation transfer (CEST) and magnetization transfer techniques provide unique and potentially quantitative contrast mechanisms in multiple MRI applications. However, the in vivo implementation of these techniques has been limited by the relatively slow MRI acquisition techniques, especially on high-field MRI scanners. A new, rapid CEST-fast imaging with steady-state free precession technique was developed to provide sensitive CEST contrast in ∼20 sec. In this study at 7 T with in vitro bovine glycogen samples and initial in vivo results in a rat liver, the CEST-fast imaging with steady-state free precession technique was shown to provide equivalent CEST sensitivity in comparison to a conventional CEST-spin echo acquisition with a 50-fold reduction in acquisition time. The sensitivity of the CEST-fast imaging with steady-state free precession technique was also shown to be dependent on k-space encoding with centric k-space encoding providing a 30-40% increase in CEST sensitivity relative to linear encoding for 256 or more k-space lines. Overall, the CEST-fast imaging with steady-state free precession acquisition technique provides a rapid and sensitive imaging platform with the potential to provide quantitative CEST and magnetization transfer imaging data.

Dual MET-EGFR combinatorial inhibition against T790M-EGFR-mediated erlotinib-resistant lung cancer.

Despite clinical approval of erlotinib, most advanced lung cancer patients are primary non-responders. Initial responders invariably develop secondary resistance, which can be accounted for by T790M-EGFR mutation in half of the relapses. We show that MET is highly expressed in lung cancer, often concomitantly with epidermal growth factor receptor (EGFR), including H1975 cell line. The erlotinib-resistant lung cancer cell line H1975, which expresses L858R/T790M-EGFR in-cis, was used to test for the effect of MET inhibition using the small molecule inhibitor SU11274. H1975 cells express wild-type MET, without genomic amplification (CNV = 1.1). At 2 microM, SU 11274 had significant in vitro pro-apoptotic effect in H1975 cells, 3.9-fold (P = 0.0015) higher than erlotinib, but had no effect on the MET and EGFR-negative H520 cells. In vivo, SU11274 also induced significant tumour cytoreduction in H1975 murine xenografts in our bioluminescence molecular imaging assay. Using small-animal microPET/MRI, SU11274 treatment was found to induce an early tumour metabolic response in H1975 tumour xenografts. MET and EGFR pathways were found to exhibit collaborative signalling with receptor cross-activation, which had different patterns between wild type (A549) and L858R/T790M-EGFR (H1975). SU11274 plus erlotinib/CL-387,785 potentiated MET inhibition of downstream cell proliferative survival signalling. Knockdown studies in H1975 cells using siRNA against MET alone, EGFR alone, or both, confirmed the enhanced downstream inhibition with dual MET-EGFR signal path inhibition. Finally, in our time-lapse video-microscopy and in vivo multimodal molecular imaging studies, dual SU11274-erlotinib concurrent treatment effectively inhibited H1975 cells with enhanced abrogation of cytoskeletal functions and complete regression of the xenograft growth. Together, our results suggest that MET-based targeted inhibition using small-molecule MET inhibitor can be a potential treatment strategy for T790M-EGFR-mediated erlotinib-resistant non-small-cell lung cancer. Furthermore, optimised inhibition may be further achieved with MET inhibition in combination with erlotinib or an irreversible EGFR-TKI.

A method for fast 3D tracking using tuned fiducial markers and a limited projection reconstruction FISP (LPR-FISP) sequence.

This work demonstrates the feasibility of using wireless, tuned fiducial markers with a limited projection reconstruction-fast imaging with steady-state free precession sequence (LPR-FISP) to accurately obtain tracking information necessary for interactive scan plane selection in magnetic resonance imaging (MRI). The position and orientation of a rigid interventional device can be uniquely determined from the 3D coordinates of three fiducial markers mounted in a known configuration on the device. Three fiducial markers were tuned to the proton resonant frequency in a 0.2T open MR scanner and mounted to the surface of a cylindrical water phantom. An LPR-FISP sequence was developed to suppress the water phantom signal while preserving that of the fiducial markers through a nonselective low-tip-angle excitation and a dephaser gradient applied prior to data acquisition. A localization algorithm was developed to accurately calculate the 3D coordinates of the fiducial markers using four LPR-FISP projections in two orthogonal scan planes. The sequence repetition time (TR = 21 msec) and the limited projection set resulted in fast LPR-FISP coordinate acquisition times of approximately 170 msec with an accuracy (max error) of 3 mm on a 0.2T MR system. This fast, accurate tracking method provides the fundamental technology for interactive MRI scan plane definition for rigid interventional devices without the need for stereotactic cameras or reference frames.