Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Prognostic impact of chronotropic incompetence in transthyretin cardiac amyloidosis: a multicentre study.

ESC heart failure·2026

Same author

Prevalence of cardiac amyloidosis in screening studies: a systematic review and meta-analysis.

ESC heart failure·2026

Same author

Functional PET imaging of gut microbiota with [<sup>18</sup>F]fluorodeoxyglucose, [<sup>18</sup>F]fluorodeoxysorbitol and [<sup>11</sup>C]choline reflects Clostridia and Lactobacillales abundance in caecum and small intestine and host metabolic interactions.

European journal of nuclear medicine and molecular imaging·2026

Same author

Mitral Valve Endocarditis by Micrococcus luteus: First European Case.

JACC. Case reports·2026

Same author

Therapy of amyloidosis: stabilizers and silencers.

European heart journal supplements : journal of the European Society of Cardiology·2026

Same author

Update on heart transplantation: a clinical consensus statement of the Advanced Heart Failure Committee of the Heart Failure Association of the European Society of Cardiology.

European journal of heart failure·2026

Search research articles

Related Experiment Video

Updated: Jul 15, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Bayesian Convolutional Neural Networks in Medical Imaging Classification: A Promising Solution for Deep Learning

Filippo Bargagna^1,2, Lisa Anita De Santi^3,4, Nicola Martini⁴

¹University of Pisa, Pisa, Italy. filippo.bargagna@phd.unipi.it.

Journal of Digital Imaging

|October 3, 2023

Summary

This summary is machine-generated.

Bayesian convolutional neural networks (BCNNs) improve medical image classification with scarce data. These models offer reliable confidence estimates and better detection of unusual inputs, enhancing diagnostic accuracy for rare diseases.

Keywords:

Bayesian convolutional neural networks Cardiac amyloidosis Data scarcity Deep learning Probabilistic programming Uncertainty

More Related Videos

DNA Virus Detection System Based on RPA-CRISPR/Cas12a-SPM and Deep Learning

DNA Virus Detection System Based on RPA-CRISPR/Cas12a-SPM and Deep Learning

Published on: May 10, 2024

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Related Experiment Videos

Last Updated: Jul 15, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

DNA Virus Detection System Based on RPA-CRISPR/Cas12a-SPM and Deep Learning

DNA Virus Detection System Based on RPA-CRISPR/Cas12a-SPM and Deep Learning

Published on: May 10, 2024

Deep Neural Networks for Image-Based Dietary Assessment

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

Area of Science:

Medical Imaging
Artificial Intelligence
Machine Learning

Background:

Deep neural networks (DNNs) excel in medical data analysis but struggle with limited datasets, leading to poor generalization and biased results.
Deterministic models lack epistemic uncertainty quantification, hindering assessment of their reliability in clinical settings.
Data scarcity is a significant challenge in medical AI, particularly for rare diseases and early-stage research.

Purpose of the Study:

To develop a probabilistic classification framework to address DNN limitations in data-scarce medical scenarios.
To implement and evaluate a Bayesian convolutional neural network (BCNN) for classifying cardiac amyloidosis (CA) subtypes.
To compare the performance and reliability of BCNNs against traditional deterministic CNNs.

Main Methods:

Four convolutional neural network (CNN) architectures were developed: base-deterministic, dropout-deterministic, dropout-Bayesian, and Bayesian.
Models were trained on a dataset of 1107 PET images from 47 patients with cardiac amyloidosis and controls, representing a data scarcity scenario.
Performance metrics included test accuracy, out-of-distribution detection, and confidence estimation.

Main Results:

The Bayesian model achieved comparable test accuracy (78.28%) to deterministic models in a data-scarce environment.
BCNNs significantly improved out-of-distribution input detection, reducing validation-test accuracy mismatch.
Dropout-Bayesian and Bayesian models provided confidence estimates, enhancing reliability and reducing failure modes of deterministic approaches.

Conclusions:

Bayesian CNNs offer a promising solution for medical image classification tasks with limited data.
The probabilistic nature of BCNNs enhances model reliability and provides crucial insights into classification confidence.
BCNNs effectively address generalization issues and silent failures common in DNNs applied to scarce medical datasets.