Unintended Consequences of Machine Learning in Medicine

By Federico Cabitza, Raffaele Rasoini, Gian Franco Gensini. 2017

472 citations recorded by the source corpus.

Subfields: General Health Professions, Health Information Management

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In General Health Professions

In Health Information Management

Citations in the corpus

Variable generalization performance of a deep learning model to detect pneumonia in chest radiographs: A cross-sectional study (2018, background): “External test data may be different in important ways from internal test data, and this may affect model performance, particularly if confounding variables exist in internal data that do not exist in external data [11].”
Impact of Deep Learning Assistance on the Histopathologic Review of Lymph Nodes for Metastatic Breast Cancer (2018, background): “23,24 In our study, pathologists were informed of the differences between the high con ﬁ dence and moderate con ﬁ dence regions, and to expect the possibility of false positives and false negatives in the assistance.”
Conversational agents in healthcare: a systematic review (2018, background): “57,58 Finally, a social-systems analysis is currently missing from research on conversational agents, an absence that has also been reported for artificial intelligence applications in previous literature.”
Potential Biases in Machine Learning Algorithms Using Electronic Health Record Data (2018, background): “8,20 Computer scientists and bioinformaticians, together with practitioners, biostatisticians, and epidemiologists, should outline the “intent behind the design,” 9(p982) including choosing appropriate questions and settings for machine learning use, interpreting findings, and conducting follow-up…”
Artificial Intelligence in Surgery: Promises and Perils (2018, background): “53 Such concerns have hindered the use of AI algorithms in many applicative fields from medicine to autonomous driving and have pushed data scientists to improve the interpretability of AI analyses.”
The ethics of AI in health care: A mapping review. (2020, background): “Consequently, an over-reliance on the ‘quantitative’ and objec-tive evidence that fuels clinical algorithms (Cabitza et al., 2017) could discredit other forms of diagnosis and treatment (Rosenfeld et al., 2019)—a prominent concern in the case of clinical psychiatry (Burns, 2015).”
Machine learning and deep learning approach for medical image analysis: diagnosis to detection (2022, background): “ML is a branch of AI where a machine learns from the data by identifying patterns and automates decision-making with minimum human intervention [96, 24, 12].”
PadChest: A large chest x-ray image dataset with multi-label annotated reports (2019, background): “Parameter Values CNN RNN Batch size 256, 512, 1024, 2048 1024 1024 L2 Penalty 0, 10−3, 10−2 0 0 Learning rate 10−4, 5 · 10−4, 10−3, 10−2 10−4 10−2 Filter size [2, 10] 3 Number of filters second conv layer 64, 128, 256 128 Dropout probability [0.”