Shorenstein Asia-Pacific Research Center’s Post

Predicting the chances of neurological recovery for cardiac arrest patients by using signal processing and machine learning on a large dataset of EEG and patient data. Check out this blog to learn more about their approach! #PhysioNetChallenge #EEGAnalysis #MachineLearning #SignalProcessing

I am happy to share that our manuscript, "An Explainable Deep Learning Approach to Classify Dementia Using Brain MRI," accepted at AICONF'24 for publication by Procedia Computer Science-Elsevier, emphasizes not only accuracy but also explainability through methods like Grad-CAM, crucial for trust in AI-enabled medical diagnostics. I thank Professor Morten Goodwin from the University of Agder (UiA) and Micheal Dutt from Egde for the guidance! #AICONF24 #DeepLearning #DementiaClassification #ExplainableAI

Just finished watching an insightful session from Samsara's AI Speaker Series session featuring Professor John Shen from CMU. Learn about the future of AI with neuromorphic computing in this video: 

An Intelligent Computer Aided Diagnosis System for Classification of Ovarian Masses using Machine Learning Approach Ovarian cancer, a difficult and often asymptomatic malignancy, remains a substantial global health concern in women. An ovary is a female reproductive organ, which lies on each side of the uterus and used to store eggs. Computer-aided diagnosis (CAD) is an approach that involves using computer algorithms and machine learning techniques to assist medical professionals in diagnosing ovarian malignancies, benign tumors or Poly-cystic ovaries (PCOS). The need for models that can effectively predict benign ovarian tumors and ovarian cancer has led to the use of machine learning techniques. Our research objective is to propose a machine learning-based system for accurate and early ovarian mass detection utilizing novel annotated ovarian masses. We have used an actual patient database whose input features were extracted from 187 transvaginal ultrasound images from database. The input image is preprocessed using the Block Matching 3D filter. The process involves employing binary and watershed segmentation techniques, followed by the integration of Gabor, Gray-Level Co-Occurrence Matrix (GLCM), Tamura, and edge feature extraction methods. K-Nearest Neighbors (KNN) and Random Forest (RF) are two classifiers used for classification. Based on our results, we are able to demonstrate that binary segmentation with RF classifiers is more accurate (above 86%) than KNN classifiers (under 84%). #ComputerAidedDiagnosis(CAD)  #TransvaginalUltrasound #BlockMatching3DFilter #BinarySegmentation #Watershed #KNN #Research #MaterialScience #Innovation #Journal #Publication #Scopus #OpenAcess #Environment #Engineering #Scientific #Science #Study #R&D Read more https://lnkd.in/gwXJ7eXt   For more information, please read our recent articles at https://lnkd.in/gqaMY-sW

As part of the 2023 George B. Moody PhysioNet Challenge, Team Swarthbeat of Swarthmore College developed an algorithm that could predict the chances of neurological recovery for cardiac arrest patients by using signal processing and machine learning on a large dataset of EEG and patient data. Check out this blog to learn more about their approach! 🧠🔬 #PhysioNetChallenge #EEGAnalysis #MachineLearning #SignalProcessing

Machine learning for seizure detection in EEG data is underway. Here's one for intraoperative neuromonitoring... With the different criteria used in tcMEP and the consequences of getting it wrong, it makes sense for this to be a first target. Change in amplitude. Change in latency. Change in turns. Change in threshold. It's all been looked at with varying degrees of success and failure. Machine learning might help give practitioners a more accurate way to interpret the data. Here's what they looked at: Test the outputs from machine learning against that of experienced professionals from the point of view of 3 different tasks. ● Task 1. Intra-patient (same patient). ● Task 2. Inter-patient (one patient against others). ● Task 3. Inter-protocol (one protocol vs the rest). This will help answer questions about what tasks are stronger than others and how they fare against the established pros. Well, there were some interesting findings (task 3 won out, ML>experts, more mm groups = harder, latencies might not be as useful as we think) I encourage you to go check out the article. There's something to learn about interpreting tcMEPs today and stay in the know about what #IONM might look like in the future. #neuromonitoring #CNIM #EEG #neurolgy Boaro, A., Azzari, A., Basaldella, F., Nunes, S., Feletti, A., Bicego, M., & Sala, F. (2024). Machine learning allows expert level classification of intraoperative motor evoked potentials during neurosurgical procedures. Computers in Biology and Medicine, 180, 109032.

Taking a #differentPerspective: Yesterday, I attend the Computing in Cardiology 2024 (#CinC2024) conference, hosted at Karlsruher Institut für Technologie (KIT) and found many common research questions between #industrialAI and #cardiacComputation: ▶ How to foster #trustworthiness in AI-based systems (e.g. with #explainableAI and #mlOps) ▶ How to improve #stability and #speed of simulations as well as #inverseProblems (e.g. with #PINNs or #GNNs) ▶ How to use AI to overcome the #mediaBreaks (e.g. using #DNNs for extracting information from paper prints) ▶ How will the #EUaiAct impact the product life-cycle? Maybe these are some leads for potential collaboration? Approach me in case of interest to discuss! And of course it was great to meet old friends again 😊 Blanca Rodriguez, Jichao Zhao, Axel Loewe, Pablo Lamata, Olaf Doessel, Pyotr Platonov, Fernando Campos, Dr. Gunnar Seemann, Martin Bishop and to discuss with current colleagues and collaborators Leonie Schicketanz and Jakub Grzelak! Computing in Cardiology #modelling #simulation #signalProcessing #AI #physicsInformedNeuralNetworks #graphNeuralNetworks #aiAct

Our review article "A Comprehensive Review on Efficient Artificial Intelligence Models for Classification of Abnormal Cardiac Rhythms using Electrocardiograms" with Utkarsh Gupta, Naveen Paluru, Deepankar Nankani and Kanchan Kulkarni got accepted in the Heliyon journal. Please find the accepted article here: https://lnkd.in/eAmA6vum #electrocardiogram #deeplearning #efficientmodels #artificialintelligence

Have questions about the brain-computer interface (BCI) landscape? If so, read on for our take on the fast-moving BCI ecosystem and what sets us apart.   ➡️ Unlike pure-play BCI companies focused on developing technology to help a paralyzed person communicate with a computer mouse, keyboard, or voice generator, we are advancing BCI technology to control our investigational implantable ARC-IM® spinal cord stimulation therapy that is designed to restore thought-driven movement of the human body after paralysis.   ➡️ The BCI we are using in clinical feasibility studies pairs ARC-IM Therapy with a BCI from Commissariat a l'Energie Atomique et aux Energies Alternatives-CLINATEC with 5 years of human safety data.   ➡️ While Neuralink is at an early clinical stage with its first human BCI implant, peer-reviewed publications have described how our ARC-IM Therapy restored movement of the legs, including the ability to walk, in 10 people with spinal cord injury and 2 with Parkinson’s disease, including one person who also received a BCI to enable more natural, thought-driven movement.    ➡️ A second person has been implanted with ARC-IM Therapy plus a BCI in a study researching thought-driven movement of the arms and hands, and at least 3 additional ARC-IM + BCI implants are planned over the next year.   ➡️ Our ARC-IM System is “BCI-ready” -- it has been designed to receive wireless signals from a BCI. That means we could eventually collaborate with Neuralink and other BCI companies.   To learn more about the unique ways in which ONWARD Medical is researching BCI technology to help people regain movement after paralysis and to put recent BCI news in context, please watch the video interview below with CEO Dave Marver. ⬇️   *All ONWARD Medical devices and therapies, including but not limited to ARC-IM®, ARC-EX®, and ARC Therapy™, alone or in combination with a brain-computer interface (BCI), are investigational and not available for commercial use. #EmpoweringMovement #SCI #BCI #ONWARDJourneys   https://lnkd.in/gW77nEFc

What is a perception ? A perceptron mimics real neurons present in our minds. Perceptrons receive inputs from several entities and apply various functions to them, transforming them into the desired output. Perceptrons primarily perform binary classifications, which means they see inputs, compute functions based on the inputs’ weight and convert the data into the required result. #DeepLearning