Computers in Biology and Medicine, 2022
Self‑supervised region‑aware segmentation of COVID‑19 CT images using 3D GAN and contrastive learning
#Segmentation #Self-supervised #COVID‑19 CT
Introduces a method combining 3D GANs with contrastive learning for self‑supervised, region‑aware segmentation of COVID‑19 lesions in CT scans.
PDFRead this if you’re curious about self-supervised segmentation in pandemic imaging.
npj Digital Medicine, 2021
CovidCTNet: an open‑source deep Learning Approach to Diagnose Covid‑19 Using Small Cohort of CT Images
#Diagnosis #CT‑classification #Covid19
Proposes an anomaly-detection network trained on synthetic CT slices to identify and classify COVID‑19 from other pneumonia types, tailored for limited data scenarios.
PDFA must-read for open-source AI in healthcare and pandemic response.
Sensors, 2020
Understanding Smartwatch Battery Utilization in the Wild
#Time‑series #IoT #BatteryAnalytics
Analyzes real-world battery usage patterns from 832 users, using clustering and a transparent convolutional neural network to distinguish high vs. low consumption events.
PDFFor those interested in wearable analytics and real-world data.
arXiv preprint, 2021
FEDZIP: A Compression Framework for Communication‑Efficient Federated Learning
#FederatedLearning #Compression #CommunicationEfficiency
Introduces FedZip, a federated-learning compression method using Top-z sparsification, clustering-based quantization, and encoding to reduce communication by up to 1000× while preserving accuracy.
PDFRead this if you care about scalable, efficient AI.
arXiv / CoRR, 2023
Beta‑Rank: A Robust Convolutional Filter Pruning Method For Imbalanced Medical Image Analysis
#Pruning #ModelCompression #MedicalImaging
Presents “Beta‑Rank,” a pruning technique designed to improve convolutional neural networks’ performance on imbalanced medical imaging datasets.
PDFCurious about model compression for real-world medical data? Start here.
27th IEEE ICEE (Iran), 2019
Mental Arousal Level Recognition Competition on the Shared Database
#AffectiveComputing #Competition #PhysiologicalSignals
Details a competition on recognizing mental arousal levels using shared physiological datasets (GSR, PPG, respiration, etc.).
For those interested in affective computing and physiological signal analysis.
ISMRM 2023 Conference
K‑space Based Motion Estimation for Polar fMRI Using Transfer Learning
#fMRI #MotionEstimation #TransferLearning
Proposes a transfer learning method for estimating motion directly from k-space data in polar fMRI scans, improving motion correction techniques.
A technical leap for fMRI motion correction.
Preprint, January 2024
PedVision: A Manual-Annotation-Free and Age Scalable Segmentation Pipeline for Bone Analysis in Hand X-Ray Images
#MedicalImaging #Segmentation #SelfSupervisedLearning #PediatricAnalysis #BoneXRay
PedVision is a segmentation pipeline designed to analyze hand X-ray images without the need for manual annotations. It is scalable across different age groups, making it suitable for pediatric bone analysis. The method utilizes self-supervised learning techniques to achieve high accuracy while minimizing the dependency on labeled data.
PDFA new era for pediatric bone imaging—no manual annotation required.
