Springer Nature has retracted 38 papers, conference proceedings and book chapters tied to a flawed autism dataset, MedPage Today reported March 19.
The dataset included more than 2,900 images of children’s faces labeled as autistic or not autistic but lacked verification of diagnoses and consent from guardians, raising ethical issues, the report said.
“From what we’ve managed to piece together, the original dataset was collected by a retired computer scientist who scraped images of children on websites related to autism, alongside a control dataset of images of children that had been scraped from across the internet,” Tim Kersjes, Springer Nature’s head of research integrity, told MedPage Today.
Spring Nature removed the affected research entirely due to the sensitive nature of the data and said most retractions are complete, with remaining actions underway, according to a spokesperson from the publishing company. The Institute of Electrical and Electronics Engineers has issued expressions of concern on 25 related articles, while other publishers, including Elsevier, PLOS and Wiley, have also taken action. Several papers using the dataset from other publications remain available with no retraction.
The dataset was removed from Kaggle — a data science and machine-learning platform owned by Google — in 2022 but later resurfaced on Google Drive, according to MedPage Today.
At the Becker's Fall Behavioral Health Summit, taking place October 20–22 in Chicago, behavioral health leaders and executives will explore strategies for expanding access to care, integrating services, addressing workforce challenges and leveraging innovation to improve outcomes across the behavioral health continuum. Apply for complimentary registration now.
