Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

oncology Oncology

Deep learning-based synthetic CT generation for paediatric brain MR-only photon and proton radiotherapy.

In Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology

Background and Purpose To enable accurate magnetic resonance imaging (MRI)-based dose calculations, synthetic computed tomography (sCT) images need to be generated. We aim at assessing the feasibility of dose calculations from MRI acquired with a heterogeneous set of imaging protocol for paediatric patients affected by brain tumours. Materials and methods Sixty paediatric patients undergoing brain radiotherapy were included. MR imaging protocols varied among patients, and data heterogeneity was maintained in train/validation/test sets. Three 2D conditional generative adversarial networks (cGANs) were trained to generate sCT from T1-weighted MRI, considering the three orthogonal planes and its combination (multi-plane sCT). For each patient, median and standard deviation (σ) of the three views were calculated, obtaining a combined sCT and a proxy for uncertainty map, respectively. The sCTs were evaluated against the planning CT in terms of image similarity and accuracy for photon and proton dose calculations.Results A mean absolute error of 61±14 HU (mean±1σ) was obtained in the intersection of the body contours between CT and sCT. The combined multi-plane sCTs performed better than sCTs from any single plane. Uncertainty maps highlighted that multi-plane sCTs differed at the body contours and air cavities. A dose difference of -0.1±0.3% and 0.1±0.4% was obtained on the D>90% of the prescribed dose and mean γ2%,2mm pass-rate of 99.5±0.8% and 99.2±1.1% for photon and proton planning, respectively. Conclusion Accurate MR-based dose calculation using a combination of three orthogonal planes for sCT generation is feasible for paediatric brain cancer patients, even when training on a heterogeneous dataset.

Maspero Matteo, Bentvelzen Laura G, Savenije Mark Hf, Guerreiro Filipa, Seravalli Enrica, Janssens Geert O, van den Berg Cornelis At, Philippens Marielle Ep


Artificial intelligence, Brain tumors, Deep learning, Image-to-image translation, Machine learning., Pediatric oncology, Radiotherapy, Synthetic CT

General General

Real-Time Assembly of Coordination Patterns in Human Infants.

In Current biology : CB

Flexibility and generativity are fundamental aspects of functional behavior that begin in infancy and improve with experience. How do infants learn to tailor their real-time solutions to variations in local conditions? On a nativist view, the developmental process begins with innate prescribed solutions, and experience elaborates on those solutions to suit variations in the body and the environment. On an emergentist view, infants begin by generating a variety of strategies indiscriminately, and experience teaches them to select solutions tailored to the current relations between their body and the environment. To disentangle these accounts, we observed coordination patterns in 11-month-old pre-walking infants with a range of cruising (moving sideways in an upright posture while holding onto a support) and crawling experience as they cruised over variable distances between two handrails they held for support. We identified infants' coordination patterns using a novel combination of computer-vision, machine-learning, and time-series analyses. As predicted by the emergentist view, the least experienced infants generated multiple coordination patterns inconsistently regardless of body size and handrail distance, whereas the most experienced infants tailored their coordination patterns to body-environment relations and switched solutions only when necessary. Moreover, the beneficial effects of experience were specific to cruising and not crawling, although both skills involve anti-phase coordination among the four limbs. Thus, findings support an emergentist view and suggest that everyday experience with the target skill may promote "learning to learn," where infants learn to assemble the appropriate solution for new problems on the fly.

Ossmy Ori, Adolph Karen E


artificial intelligence, behavioral flexibility, computer vision, cruising, infants, limb coordination, locomotion, machine learning, motor development, problem solving

General General

Targeting thermoTRP ion channels: in silico preclinical approaches and opportunities.

In Expert opinion on therapeutic targets

INTRODUCTION : A myriad of cellular pathophysiological responses are mediated by polymodal ion channels that respond to chemical and physical stimuli such as thermoTRP channels. Intriguingly, these channels are pivotal therapeutic targets with limited clinical pharmacology. In silico methods offer an unprecedented opportunity for discovering new lead compounds targeting thermoTRP channels with improved pharmacological activity and therapeutic index.

AREAS COVERED : This article reviews the progress on thermoTRP channel pharmacology because of (i) advances in solving their atomic structure using cryo-electron microscopy and, (ii) progress on computational techniques including homology modeling, molecular docking, virtual screening, molecular dynamics, ADME/Tox and artificial intelligence. Together, they have increased the number of lead compounds with clinical potential to treat a variety of pathologies. We used original and review articles from Pubmed (1997-2020), as well as the database, containing the terms thermoTRP, artificial intelligence, docking, and molecular dynamics.

EXPERT OPINION : The atomic structure of thermoTRP channels along with computational methods constitute a realistic first line strategy for designing drug candidates with improved pharmacology and clinical translation. In silico approaches can also help predict potential side-effects that can limit clinical development of drug candidates. Together, they should provide drug candidates with upgraded therapeutic properties.

Fernández-Ballester Gregorio, Fernández-Carvajal Asia, Ferrer-Montiel Antonio


ADME, artificial intelligence, docking, ion channel, molecular dynamics, thermoTRP channels, virtual screening

General General

Next-Generation Analytics for Omics Data.

In Cancer cell ; h5-index 124.0

The increasing omics data present a daunting informatics challenge. DrBioRight, a natural language-oriented and artificial intelligence-driven analytics platform, enables the broad research community to perform analysis in an intuitive, efficient, transparent, and collaborative way. The emerging next-generation analytics will maximize the utility of omics data and lead to a new paradigm for biomedical research.

Li Jun, Chen Hu, Wang Yumeng, Chen Mei-Ju May, Liang Han


General General

Artificial Intelligence-based detection of lymph node metastases by PET/CT predicts prostate cancer-specific survival.

In Clinical physiology and functional imaging

INTRODUCTION : Lymph node metastases are a key prognostic factor in prostate cancer (PCa), but detecting lymph node lesions from PET/CT images is a subjective process resulting in inter-reader variability. Artificial intelligence (AI)-based methods can provide an objective image analysis. We aimed at developing and validating an AI-based tool for detection of lymph node lesions.

METHODS : A group of 399 patients with biopsy-proven PCa who had undergone 18 F-choline PET/CT for staging prior to treatment were used to train (n=319) and test (n=80) the AI-based tool. The tool consisted of convolutional neural networks using complete PET/CT scans as inputs. In the test set, the AI-based lymph node detections were compared to those of two independent readers. The association with PCa-specific survival was investigated.

RESULTS : The AI-based tool detected more lymph node lesions than Reader B (98 vs 87/117; p=0.045) using Reader A as reference. AI-based tool and Reader A showed similar performance (90 vs 87/111; p=0.63) using Reader B as reference. The number of lymph node lesions detected by the AI-based tool, PSA, and curative treatment were significantly associated with PCa-specific survival.

CONCLUSION : This study shows the feasibility of using an AI-based tool for automated and objective interpretation of PET/CT images that can provide assessments of lymph node lesions comparable with that of experienced readers, and prognostic information in PCa patients.

Borrelli Pablo, Larsson Måns, Ulén Johannes, Enqvist Olof, Trägårdh Elin, Hvid Poulsen Mads, Mortensen Mike Allan, Kjölhede Henrik, Høilund-Carlsen Poul Flemming, Edenbrandt Lars


Artificial intelligence, Fluorocholine, Lymph node metastases, PCa, PET

General General

Spage2vec: Unsupervised representation of localized spatial gene expression signatures.

In The FEBS journal

Investigations of spatial cellular composition of tissue architectures revealed by multiplexed in situ RNA detection often rely on inaccurate cell segmentation or prior biological knowledge from complementary single cell sequencing experiments. Here we present spage2vec, an unsupervised segmentation free approach for decrypting the spatial transcriptomic heterogeneity of complex tissues at subcellular resolution. Spage2vec represents the spatial transcriptomic landscape of tissue samples as a graph and leverages a powerful machine learning graph representation technique to create a lower dimensional representation of local spatial gene expression. We apply spage2vec to mouse brain data from three different in situ transcriptomic assays and to a spatial gene expression dataset consisting of hundreds of individual cells. We show that learned representations encode meaningful biological spatial information of re-occuring localized gene expression signatures involved in cellular and subcellular processes.

Partel Gabriele, Wählby Carolina


RNA profiling, Spatial transcriptomics, gene expression, graph representation learning, tissue analysis