Human–machine partnership with artificial intelligence for chest radiograph diagnosis

Abstract

Human-in-the-loop (HITL) AI may enable an ideal symbiosis of human experts and AI models, harnessing the advantages of both while at the same time overcoming their respective limitations. The purpose of this study was to investigate a novel collective intelligence technology designed to amplify the diagnostic accuracy of networked human groups by forming real-time systems modeled on biological swarms. Using small groups of radiologists, the swarm-based technology was applied to the diagnosis of pneumonia on chest radiographs and compared against human experts alone, as well as two state-of-the-art deep learning AI models. Our work demonstrates that both the swarm-based technology and deep-learning technology achieved superior diagnostic accuracy than the human experts alone. Our work further demonstrates that when used in combination, the swarm-based technology and deep-learning technology outperformed either method alone. The superior diagnostic accuracy of the combined HITL AI solution compared to radiologists and AI alone has broad implications for the surging clinical AI deployment and implementation strategies in future practice.

Assessment of dysplasia in bone marrow smear with convolutional neural network

“…This indicates that AKIRA complemented human labelling and thereby enhanced its own accuracy. Such 'human in the loop' iteration methods are effective for AI systems that generate training data based on human judgement, as previously reported by another group 24 .…”

Artificial intelligence and the ongoing need for empathy, compassion and trust in healthcare

“…29 Furthermore, human-machine partnerships can provide far better results than either humans or machines alone. 30 In these examples, the principal benefits of artificial intelligence stem from its ability to improve efficiency and effectiveness by guiding diagnoses, delivering more accurate results and thus eliminating human error. With regard to greater efficiency through prevention, artificial intelligence technologies that track and analyse the movement of individuals could be used to detect people at risk of stroke and eliminate that risk through early intervention.…”

TzanckNet: a convolutional neural network to identify cells in the cytology of erosive-vesiculobullous diseases

“…We would like to emphasize that we propose this model as a clinical decision support system to improve physician accuracy and efficiency in the clinical workflow, not as a replacement or a tool that can directly output a diagnosis. It has been shown that human–machine combination works better than either alone 39 – 41 . While it is possible to make a machine learning model that can directly output a diagnosis, we consider such black box approaches to be unsuitable for high stakes decisions like medical diagnosis.…”