Multiple instance learning framework can facilitate explainability in murmur detection.

OBJECTIVE: Cardiovascular diseases (CVDs) account for a high fatality rate worldwide. Heart murmurs can be detected from phonocardiograms (PCGs) and may indicate CVDs. Still, they are often overlooked as their detection and correct clinical interpretation require expert skills. In this work, we aim to predict the presence of murmurs and clinical outcomes from multiple PCG recordings employing an explainable multitask model. APPROACH: Our approach consists of a two-stage multitask model. In the first stage, we predict the murmur presence in single PCGs using a multiple instance learning (MIL) framework. MIL also allows us to derive sample-wise classifications (i.e. murmur locations) while only needing one annotation per recording ("weak label") during training. In the second stage, we fuse explainable hand-crafted features with features from a pooling-based artificial neural network (PANN) derived from the MIL framework. Finally, we predict the presence of murmurs and the clinical outcome for a single patient based on multiple recordings using a simple feed-forward neural network. MAIN RESULTS: We show qualitatively and quantitatively that the MIL approach yields useful features and can be used to detect murmurs on multiple time instances and may thus guide a practitioner through PCGs. We analyze the second stage of the model in terms of murmur classification and clinical outcome. We achieved a weighted accuracy of 0.714 and an outcome cost of 13612 when using the PANN model and demographic features on the CirCor dataset (hidden test set of the George B. Moody PhysioNet challenge 2022, team "Heart2Beat", rank 12 / 40). SIGNIFICANCE: To the best of our knowledge, we are the first to demonstrate the usefulness of MIL in PCG classification. Also, we showcase how the explainability of the model can be analyzed quantitatively, thus avoiding confirmation bias inherent to many post-hoc methods. Finally, our overall results demonstrate the merit of employing MIL combined with handcrafted features for the generation of explainable features as well as for a competitive classification performance.

Automatic Generation of Labeled Data for Video-Based Human Pose Analysis via NLP applied to YouTube Subtitles.

With recent advancements in computer vision as well as machine learning (ML), video-based at-home exercise evaluation systems have become a popular topic of current research. However, performance depends heavily on the amount of available training data. Since labeled datasets specific to exercising are rare, we propose a method that makes use of the abundance of fitness videos available online. Specifically, we utilize the advantage that videos often not only show the exercises, but also provide language as an additional source of information. With push-ups as an example, we show that through the analysis of subtitle data using natural language processing (NLP), it is possible to create a labeled (irrelevant, relevant correct, relevant incorrect) dataset containing relevant information for pose analysis. In particular, we show that irrelevant clips (n = 332) have significantly different joint visibility values compared to relevant clips (n = 298). Inspecting cluster centroids also show different poses for the different classes.

Video-Based Hand Movement Analysis of Parkinson Patients before and after Medication Using High-Frame-Rate Videos and MediaPipe.

Tremor is one of the common symptoms of Parkinson's disease (PD). Thanks to the recent evolution of digital technologies, monitoring of PD patients' hand movements employing contactless methods gained momentum. Objective: We aimed to quantitatively assess hand movements in patients suffering from PD using the artificial intelligence (AI)-based hand-tracking technologies of MediaPipe. Method: High-frame-rate videos and accelerometer data were recorded from 11 PD patients, two of whom showed classical Parkinsonian-type tremor. In the OFF-state and 30 Minutes after taking their standard oral medication (ON-state), video recordings were obtained. First, we investigated the frequency and amplitude relationship between the video and accelerometer data. Then, we focused on quantifying the effect of taking standard oral treatments. Results: The data extracted from the video correlated well with the accelerometer-based measurement system. Our video-based approach identified the tremor frequency with a small error rate (mean absolute error 0.229 (±0.174) Hz) and an amplitude with a high correlation. The frequency and amplitude of the hand movement before and after medication in PD patients undergoing medication differ. PD Patients experienced a decrease in the mean value for frequency from 2.012 (±1.385) Hz to 1.526 (±1.007) Hz and in the mean value for amplitude from 8.167 (±15.687) a.u. to 4.033 (±5.671) a.u. Conclusions: Our work achieved an automatic estimation of the movement frequency, including the tremor frequency with a low error rate, and to the best of our knowledge, this is the first paper that presents automated tremor analysis before/after medication in PD, in particular using high-frame-rate video data.