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

General General

Available Sensory Input Determines Motor Performance and Strategy in Early Blind and Sighted Short-Tailed Opossums.

In iScience

The early loss of vision results in a reorganized neocortex, affecting areas of the brain that process both the spared and lost senses, and leads to heightened abilities on discrimination tasks involving the spared senses. Here, we used performance measures and machine learning algorithms that quantify behavioral strategy to determine if and how early vision loss alters adaptive sensorimotor behavior. We tested opossums on a motor task involving somatosensation and found that early blind animals had increased limb placement accuracy compared with sighted controls, while showing similarities in crossing strategy. However, increased reliance on tactile inputs in early blind animals resulted in greater deficits in limb placement and behavioral flexibility when the whiskers were trimmed. These data show that compensatory cross-modal plasticity extends beyond sensory discrimination tasks to motor tasks involving the spared senses and highlights the importance of whiskers in guiding forelimb control.

Englund Mackenzie, Faridjoo Samaan, Iyer Christopher S, Krubitzer Leah


Behavioral Neuroscience, Biological Sciences, Neuroscience, Sensory Neuroscience

General General

MitoSegNet: Easy-to-use Deep Learning Segmentation for Analyzing Mitochondrial Morphology.

In iScience

While the analysis of mitochondrial morphology has emerged as a key tool in the study of mitochondrial function, efficient quantification of mitochondrial microscopy images presents a challenging task and bottleneck for statistically robust conclusions. Here, we present Mitochondrial Segmentation Network (MitoSegNet), a pretrained deep learning segmentation model that enables researchers to easily exploit the power of deep learning for the quantification of mitochondrial morphology. We tested the performance of MitoSegNet against three feature-based segmentation algorithms and the machine-learning segmentation tool Ilastik. MitoSegNet outperformed all other methods in both pixelwise and morphological segmentation accuracy. We successfully applied MitoSegNet to unseen fluorescence microscopy images of mitoGFP expressing mitochondria in wild-type and catp-6ATP13A2 mutant C. elegans adults. Additionally, MitoSegNet was capable of accurately segmenting mitochondria in HeLa cells treated with fragmentation inducing reagents. We provide MitoSegNet in a toolbox for Windows and Linux operating systems that combines segmentation with morphological analysis.

Fischer Christian A, Besora-Casals Laura, Rolland Stéphane G, Haeussler Simon, Singh Kritarth, Duchen Michael, Conradt Barbara, Marr Carsten


Artificial Intelligence, Automation in Bioinformatics, Bioinformatics, Cell Biology

Radiology Radiology

Fusion of medical imaging and electronic health records using deep learning: a systematic review and implementation guidelines.

In NPJ digital medicine

Advancements in deep learning techniques carry the potential to make significant contributions to healthcare, particularly in fields that utilize medical imaging for diagnosis, prognosis, and treatment decisions. The current state-of-the-art deep learning models for radiology applications consider only pixel-value information without data informing clinical context. Yet in practice, pertinent and accurate non-imaging data based on the clinical history and laboratory data enable physicians to interpret imaging findings in the appropriate clinical context, leading to a higher diagnostic accuracy, informative clinical decision making, and improved patient outcomes. To achieve a similar goal using deep learning, medical imaging pixel-based models must also achieve the capability to process contextual data from electronic health records (EHR) in addition to pixel data. In this paper, we describe different data fusion techniques that can be applied to combine medical imaging with EHR, and systematically review medical data fusion literature published between 2012 and 2020. We conducted a systematic search on PubMed and Scopus for original research articles leveraging deep learning for fusion of multimodality data. In total, we screened 985 studies and extracted data from 17 papers. By means of this systematic review, we present current knowledge, summarize important results and provide implementation guidelines to serve as a reference for researchers interested in the application of multimodal fusion in medical imaging.

Huang Shih-Cheng, Pareek Anuj, Seyyedi Saeed, Banerjee Imon, Lungren Matthew P


Data integration, Machine learning, Medical imaging

General General

Ensemble learning predicts multiple sclerosis disease course in the SUMMIT study.

In NPJ digital medicine

The rate of disability accumulation varies across multiple sclerosis (MS) patients. Machine learning techniques may offer more powerful means to predict disease course in MS patients. In our study, 724 patients from the Comprehensive Longitudinal Investigation in MS at Brigham and Women's Hospital (CLIMB study) and 400 patients from the EPIC dataset, University of California, San Francisco, were included in the analysis. The primary outcome was an increase in Expanded Disability Status Scale (EDSS) ≥ 1.5 (worsening) or not (non-worsening) at up to 5 years after the baseline visit. Classification models were built using the CLIMB dataset with patients' clinical and MRI longitudinal observations in first 2 years, and further validated using the EPIC dataset. We compared the performance of three popular machine learning algorithms (SVM, Logistic Regression, and Random Forest) and three ensemble learning approaches (XGBoost, LightGBM, and a Meta-learner L). A "threshold" was established to trade-off the performance between the two classes. Predictive features were identified and compared among different models. Machine learning models achieved 0.79 and 0.83 AUC scores for the CLIMB and EPIC datasets, respectively, shortly after disease onset. Ensemble learning methods were more effective and robust compared to standalone algorithms. Two ensemble models, XGBoost and LightGBM were superior to the other four models evaluated in our study. Of variables evaluated, EDSS, Pyramidal Function, and Ambulatory Index were the top common predictors in forecasting the MS disease course. Machine learning techniques, in particular ensemble methods offer increased accuracy for the prediction of MS disease course.

Zhao Yijun, Wang Tong, Bove Riley, Cree Bruce, Henry Roland, Lokhande Hrishikesh, Polgar-Turcsanyi Mariann, Anderson Mark, Bakshi Rohit, Weiner Howard L, Chitnis Tanuja


Multiple sclerosis

Cardiology Cardiology

Best practices for authors of healthcare-related artificial intelligence manuscripts.

In NPJ digital medicine

Since its inception in 2017, npj Digital Medicine has attracted a disproportionate number of manuscripts reporting on uses of artificial intelligence. This field has matured rapidly in the past several years. There was initial fascination with the algorithms themselves (machine learning, deep learning, convoluted neural networks) and the use of these algorithms to make predictions that often surpassed prevailing benchmarks. As the discipline has matured, individuals have called attention to aberrancies in the output of these algorithms. In particular, criticisms have been widely circulated that algorithmically developed models may have limited generalizability due to overfitting to the training data and may systematically perpetuate various forms of biases inherent in the training data, including race, gender, age, and health state or fitness level (Challen et al. BMJ Qual. Saf. 28:231-237, 2019; O'neil. Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy, Broadway Book, 2016). Given our interest in publishing the highest quality papers and the growing volume of submissions using AI algorithms, we offer a list of criteria that authors should consider before submitting papers to npj Digital Medicine.

Kakarmath Sujay, Esteva Andre, Arnaout Rima, Harvey Hugh, Kumar Santosh, Muse Evan, Dong Feng, Wedlund Leia, Kvedar Joseph


Diagnosis, Outcomes research, Publication characteristics

General General

Precise Correlation of Contact Area and Forces in the Unstable Friction between a Rough Fluoroelastomer Surface and Borosilicate Glass.

In Materials (Basel, Switzerland)

Stick-slip friction of elastomers arises due to adhesion, high local strains, surface features, and viscous dissipation. In situ techniques connecting the real contact area to interfacial forces can reveal the contact evolution of a rough elastomer surface leading up to gross slip, as well as provide high-resolution dynamic contact areas for improving current slip models. Samples with rough surfaces were produced by the same manufacturing processes as machined seals. In this work, a machined fluoroelastomer (FKM) hemisphere was slid against glass, and the stick-slip behavior was captured optically in situ. The influence of sliding velocity on sliding behavior was studied over a range of speeds from 1 µm/s to 100 µm/s. The real contact area was measured from image sequences thresholded using Otsu's method. The motion of the pinned region was delineated with a machine learning scheme. The first result is that, within the macroscale sticking, or pinned phase, local pinned and partial slip regions were observed and modeled as a combined contact with contributions to friction by both regions. As a second result, we identified a critical velocity below which the stick-slip motion converted from high frequency with low amplitude to low frequency with high amplitude. This study on the sliding behavior of a viscoelastic machined elastomer demonstrates a multi-technique approach which reveals precise changes in contact area before and during pinning and slip.

Wang Chao, Bonyadi Shabnam Z, Grün Florian, Pinter Gerald, Hausberger Andreas, Dunn Alison C


elastomer stick-slip, in-situ microtribometry, machined seals